First Solar, Inc. (FSLR) Q3 2012 Earnings Report, Transcript and Summary

Good day, everyone, and welcome to First Solar's Third Quarter 2012 Earnings Call. This call is being webcast live on the investors section of First Solar's website at firstsolar.com. [Operator Instructions] As a reminder, today's call is being recorded. I would now like to turn the call over to Mr. David Brady, Vice President of Treasury and Investor Relations for First Solar Inc. Mr. Brady, you may begin, sir.

Thank you, Joyce. Good afternoon, everyone, and thank you for joining us. Today, the company issued a press release announcing its financial results for the third quarter of 2012. A copy of the press release and the presentation are available on the Investors section of First Solar's website at firstsolar.com. With me today are Jim Hughes, Chief Executive Officer; and Mark Widmar, Chief Financial Officer. Jim will outline many of the developments we've made in sustainable markets and provide update from our forecast in volume of business and some of our technological advancements. Mark will review third quarter financial results and update guidance for 2012. We will then open up the call for questions. Most of the financial numbers reported and discussed on today's call are based on U.S. Generally Accepted Accounting Principles. In the few cases where we report non-GAAP measures, we provide a reconciliation to GAAP equivalent at the back of our presentation. Please note that during the course of this call, the company will make projections and other comments that are forward-looking statements within the meaning of the federal securities laws. The forward-looking statements in this call are based on current information and expectations, are subject to uncertainties and changes in circumstances and do not constitute guarantees of future performance. Those statements involve a number of factors that could cause actual results to differ materially from those statements, including the risks as described in the company's most recent annual report on Form 10-K and other filings with the Securities and Exchange Commission. First Solar assumes no obligation to update any forward-looking information contained in this call or with respect to the announcements described herein. It is now my pleasure to introduce Jim Hughes, Chief Executive Officer. Jim?

Thanks, David, and welcome to our third quarter 2012 earnings call. First off, I would like to apologize for any inconvenience caused by the delay in this call from yesterday to today. Given the broad impact on the Atlantic coast area and our own need to assess any impacts on First Solar operations as a result of Hurricane Sandy, we felt the delay was prudent. Secondly, I would like to extend the sympathies and support of the First Solar team to all of those individuals and families impacted by the hurricane. As a long time resident of the Texas Gulf Coast prior to joining First Solar, I well understand the vast devastation that such a storm can make. We have a number of facilities in operations in the impacted area and fortunately, we do not believe that any of our personnel suffered any injuries. Our Bridgewater, New Jersey engineering center has reopened for business with no damage. Unfortunately, we do have several construction sites impacted and we are assessing any potential damage and related impacts on the projects. At this time, we believe the impact will primarily be timing related and will not involve any material loss for the company. Mark will cover this later in this review of our financial results and guidance. As has been the case for the last several earnings calls, the solar industry remains challenged. We have seen an increasing number of production cuts, capacity reductions and bankruptcies. However, we do see signs of stabilization and a generally better feel to the marketplace although this is largely anecdotal and intangible in nature today. In particular, developments in the sustainable markets that we are targeting continue to be very encouraging and together with good levels of continuing activity in our traditional markets has allowed the company to be in a solid position financially with an expanding global presence. For example, we recently announced another significant milestone for First Solar, our first major project in the Middle East. Following a competitive bidding process, First Solar was selected by the Dubai Electricity & Water Authority to construct a 13 megawatts PV power plant near, Dubai providing both modules and EPC services. This is the first space of the landmark Mohammed bin Rashid Al Maktoum Solar Park named after the leader of the emirate of Dubai, a $3.3 billion equivalent project that is expected to eventually represents 1 gigawatt of clean energy capacity, using both PV and solar thermal technology. This is especially significant because our Cattail Technology was selected for this high-profile project after a rigorous technical review due to its superior performance and extreme desert conditions. It also illustrates our early success in executing our strategy to provide comprehensive solar power plant solutions in sustainable markets and serve rapidly growing fundamental power needs. In addition, we recently opened an office in Dubai and are in the process of establishing an office in Saudi Arabia as well. We are actively bidding on other projects in the region. In India, we announced 75 megawatts DC of new deals in the third quarter, a project with Kiran Energy Solar Power and Mahindra Solar One for a combined 50 megawatts and the 25-megawatt deal with Green Infra announced in September which are both located in Rajasthan. Year-to-date, our total of new bookings in India is 114 megawatts DC. In China, we have recently hired Bruce Yang as Managing Director and Vice President of Business Development. Bruce, formerly of British Petroleum has 25 years of experience in the energy industry throughout Asia and Europe spanning oil, gas, coal power and renewable energy. Bruce's hire is evidence of our continued commitment to the Chinese market beginning the development of the Ordos Project which Bruce will be focusing on from the outset. In the Asia Pacific region, we continue to build momentum, following the announcement of the AGL deal in Australia last quarter. We have just signed and MoU with the PJB Services of Indonesia to collaborate on the delivery of 100 megawatts of utility scale, solar power plants to address the growing energy demand in that country. PJB Services is a leading provider of services for the operation and maintenance of conventional power plants in Indonesia and this MoU represents its first foray into the development of utility scale solar PV plants. This agreement underscores our belief that the Indonesian market has great potential as a sustainable market where solar power can be a meaningful part of the energy mix. In addition, we have established a Thailand operating subsidiary to further develop that market. There are already 12 megawatts of solar PV projects in Thailand using First Solar modules. Switching to the U.S., we added power purchase agreement with PG&E for the Lost Hills and Cuyama projects we're currently developing in Central California, representing a combined 72 megawatts AC. The PPAs are subject to approval by the California Public Utilities Commission. This decision is expected in the first half of 2013. We also announced agreements to provide our EPC services and modules for projects totaling 20 megawatts AC of generating capacity for PNM Resources in New Mexico. These plants could be in service by the end of 2013, subject to approval by the New Mexico Public Regulation Commission, whose decision is expected this quarter. In addition, we have a previously unannounced agreement to supply 267 megawatts DC of our modules to a new project in California, of which we previously shipped 40 megawatts. We will provide details on the project at a later date, but we expect to ship the remaining modules in 2013. Year-to-date, new contracted orders booked for system EPC and third-party module sales in the U.S. have been 644 megawatts DC. With the addition of the DEWA project, our project pipeline now stands at 3 gigawatts AC, which represents the sum of the contracted megawatts of the projects in our pipeline, excluding those that have been delivered through Q3 2012. As of the end of the third quarter, we have recognized revenue for approximately 626 megawatts AC equivalents. Slide #7 shows the total estimated revenue associated with future business and the change year-to-date in that revenue. The data on this slide differs to the prior project pipeline slide in a number of ways. This data represents our total business which includes third-party module sales in addition to the contracted and non-contracted portions of our project pipeline. The prior slide showed the project volume on a gross basis. This data reflects the expected net revenue of total business after revenue that has been recognized during the course of the year and adding new bookings over the same period. This is a more robust reflection of the estimated total revenue of business that lies ahead of us. We have close to $9 billion of expected revenue in the future. That is just $500 million lower than at the start of the year because although we recognized $2.3 billion of revenue during that time, we compensated for that by adding $1.8 billion of new bookings at the same time. This data is as of the end of the third quarter, but we have since added more than $180 million in new bookings since then. This gives us a solid foundation to weather the current challenges facing the industry as we continue our strategy to develop the tremendous long-term growth potential of sustainable markets. On October 2, it we announced that Sumitomo Corporation purchased from General Electric 25% of the equity in the Desert Sunlight project we are constructing in California. Having a large international investor like Sumitomo take a significant stake in a U.S. solar project this size that we built is further validation of our plant reliability, services and technology globally. Meanwhile, GE remains a valuable business partner and will continue to consider First Solar projects for their renewable investment portfolio in the future. Lastly, we continue to work on active bids, project development, establishing partnerships, developing our fully-integrated capabilities and exploring acquisitions in each of our target markets around the world including Chile, Australia, North and South Africa, to name a few in addition to those already mentioned above. We are very proud of the traction and success we have achieved in these new markets in such a short space of time and we plan to continue to deliver results to validate our strategy to develop sustainable markets going forward. Turning to Slide 8, we recently announced the expansion of our Network Operations Center in Mesa, Arizona, which allows us to optimize our customers power plants to produce the maximum amount of energy and revenue under their power purchase agreements while minimizing cost and risks. It combines power prediction and analytical capabilities with advanced diagnostics. Using First Solar's advanced plant controls, the world’s largest solar PV power plants integrate seamlessly with the electrical grid and contribute to grid stability. Grid stability is important because if a solar plant or any power generation asset does not provide power into the electricity grid within certain parameters, namely within certain frequency and voltage ranges, it risks going offline and therefore lose valuable revenue. Alternatively, without such grid stability, a solar plant could create grid volatility, requiring increased support from other generating assets. Thus, the owner of the solar plant could cause additional cost to the utilities for the higher cost generating assets that had to be powered up to compensate for those assets that went offline. Either way, it can negatively impact the solar plant provider in a variety of ways. Lost revenue on the plant hurts the buyer of that plant, reducing the attractiveness of future plants to potential investors. It increases the volatility for the grid operator and therefore, the cost to utilities, making solar plants less attractive as a reliable source of power to the users of the power and therefore, once again, making such plants less attractive to future investors. Features of our grid stability technology include ramp rate control which limits how fast the power plants output increases or decreases in order to minimize gird disruption, ride through capability which enables the power plant to operate through faults and other grid disturbances, active power control which can be used to modulate power output and frequency droop control which enables the power plant to provide critical grid support when grid frequency is changing. These tools have become increasingly important as more solar generating capacity is connected to the group. Many of our competitors do not incorporate such grid stability technology into their systems, making them less predictable and reliable from a power output and rate of return perspective. First Solar has invested considerable time and resources developing this technology, making it both difficult and expensive to replicate. In order to continue our progress in the markets we are targeting, we need to be prepared to price solar electricity at levels sufficient to generate new demand without the benefit of subsidies. The principal driver of our cost roadmap is conversion efficiency but more importantly, to lower our LCOE to the targeted range of $100 to $140 per megawatt hour. We have invested heavily in our R&D team to drive genuine improvements in our technology to meet or beat our existing roadmap, and I am happy to report that we are seeing significant progress in this regard. One, we have already partially rolled our changes to our laser scribe process that result in improvement to our module active area loss. We expect to complete the rollout by mid-2013. Improvements in our semiconductor absorber material through modifications to our deposition technology are expected to go into full volume production beginning in mid-2013. We are also working on a module design change that will reduce active area losses and improve uniformity which is also expected to be rolled out in the second half of 2013. Lastly and perhaps most significant in terms of the resultant increase in efficiency, are improvements to the back contact of the module. Our current technology roadmap targets the second half of 2014 to start high-volume manufacturing rollout. Please note that these improvements all contribute to increased module conversion efficiency, reduce the so-called balance of systems penalty, are based on genuine technology improvements as opposed to lower input costs and are just a portion of the ongoing research that we're working on. Research that will not only ensure we achieve our existing cost roadmaps, but could accelerate their delivery and lower the overall cost targets. This is very encouraging in an extremely competitive industry where the lowest cost providers with the most comprehensive and customized solutions will ultimately be the winners. In summary, the solar market remains challenging but we are continuing to gain traction in the new sustainable markets we're targeting and expanding our global presence. We continue to add to our pipeline of contracted volume which includes one of the largest captive solar pipelines in the world. Our plants continue to be sought after by the top independent power producers and financial investors in the industry. We continue to execute on our cost and efficiency roadmaps. And with the significant improvements our R&D team is delivering, we believe we will not only maintain, but increase our cost advantages at the system level relative to our peers for the foreseeable future. Now, I'd like to hand the call over to Mark, who will discuss our Q3 financial results and update our 2012 financial guidance.

Okay. Thanks, Jim, and good afternoon. Starting with operations on Slide 11. In Q3, we produced 490 megawatts, up 33% from Q2. This equates to running our factories at 83% of their capacity, up from 63% in Q2. As mentioned in the second quarter call, this plant increase was in response to stronger demand, in addition the high-utilization rate was partially driven by the benefit of completed upgrades on various lines in Perrysburg and Malaysia. During Q4, we plan to run our plants at approximately 90% to 95% capacity utilization, thus continuing to minimize underutilization headwinds. Note as previously announced, the Frankfurt Oder manufacturing facility will continue production until the end of this year and then it will be shut down permanently. Model manufacturing cost per watt in the third quarter, excluding our German plant, was $0.67. On a comparable basis, module manufacturing cost per watt was down $0.05 quarter-over-quarter. This sequential decrease is primarily due to the lower plant underutilization cost and higher module efficiency. Had our plants operated at a full production for the entire quarter, our module manufacturing cost per watt would have been $0.64. Assuming full utilization under our best plant is manufacturing modules at a cost of $0.62 per plant, which on the year-on-year basis is $0.10 lower. As we exit the year, our consolidated manufacturing cost and our best line cost per watt are forecasted to be $0.64 and $0.61, respectively. This is $0.01 lower than our previous estimate for our consolidated cost per watt and $0.02 better than our previous estimate for our best line cost per watt. The average line conversion efficiency of our module was 12.7% in the third quarter, which is up 0.9 percentage points year-over-year and up 0.1 percentage point quarter-over-quarter. We achieved this average efficiency 1 quarter sooner than planned. As of the end of October, our best line is currently producing an average module efficiency of 13.2%. When this module efficiency gets combined with our lowest cost manufacturing plant, the module cost per watt would be $0.59. We believe this best represents the current operational entitlement of our module cost profile. As we assimilate learnings and assuming full utilization, from our best cost line and our best efficiency line across our manufacturing portfolio, we expect to drive consolidated manufacturing cost per watt to below $0.60 in late 2013. Note as previously communicated, this cost includes warranty, freight and end-of-life costs associated with the module. Slide 12 highlights the meaningful cost reduction and efficiency improvements, we continue to make year-over-year from both the consolidated perspective as well as the best plant. Our full year 2012 forecasted cost per watt, excluding Frankfurt Oder, is expected to between $0.68 and $0.70 per watt. Our best plant between $0.63 and $0.64. Regarding efficiency, we expect to exit 2012 at a consolidated efficiency of 12.8%. Moving on to the P&L portion of the presentation on Slide 13. Net sales for the third quarter were $839 million, down from $957 million last quarter. The decrease in net sales was primarily due to project-specific decreases, including Silver State North which was completed in the second quarter and reduced construction activity at Agua Caliente consistent with the company's planned construction schedule. The decrease was partially offset by initial revenue recognition for Topaz which began in late 2011. Initial revenue recognition was met in Q3, one quarter earlier than had planned. But the Topaz forecast for the full year is unchanged. However, sequentially, we anticipate Topaz's revenue in Q4 to be higher than the Q3 revenue. Our solar power systems revenue, which includes both our EPC revenue and solar modules used in the systems business, increased from 86% of sales in the second quarter to 95% of sales in the third quarter. Gross margin in Q3 was 28.4%, up 2.9 percentage points from the prior quarter. The gross margin improvement is mainly reflective of higher production volumes, resulting in lower underutilization charges and a favorable systems business project mix. Third quarter gross margin included $15.8 million in product warranty charges in excess of our normal warranty accrual. More specifically to this charge, as disclosed in early October, we identified a workmanship issue affecting a limited number of solar modules manufactured from October 2008 to June of 2009. A small portion of the modules manufactured during that time used a new adhesive material and process to attach the core plate or junction box to the module, which we determined may not adhere securely over time. We know the serial numbers of the affected modules and are proactively contacting system owners to repair or replace the 232,000 affected modules in a manner consistent with our workmanship warranty. For roof-mounted systems, we will also remove and replace the affected modules at no cost to the system owner, which is above and beyond our standard workmanship warranty. Operating expenses were up $28 million quarter-over-quarter to $132 million as Q2 benefited from a stock-based compensation credit associated with a change in our estimated forfeiture rate per share-based compensation awards. Q3 operating expenses were also unfavorably impacted by a onetime impairment charge of $4 million related to certain IT systems, and restructuring charges of $24.2 million. We expect fourth quarter operating expenses, excluding restructuring charges, to be approximately $102 million and the end of year normalized exit rate to be less than $100 million per quarter, excluding start-ups. When we announced our restructuring, we said that we anticipated up to $510 million of related charges. Year-to-date, we have incurred $440 million of restructuring charges. And for the fourth quarter of 2012, we expect additional restructuring charges of $20 million to $25 million. This will result in a full year charge consistent with our April 17 announcement. The balance of the total anticipated program charges are estimated to be incurred in 2013. We expect these initiatives to reduce our ongoing annualized costs by between $100 million to $120 million, split approximately equally between cost of goods sold and SG&A. On a reported basis, the third quarter 2012 operating income was $107 million compared to operating income of $140 million in the second quarter of 2012. Excluding restructuring and manufacturing excursion costs in excess of normal warranty, our second quarter operating income was $172 million. And the third quarter operating income, excluding restructuring, was $130 million. The operating income decline was primarily reflective of the lower revenue in Q3. The third quarter GAAP net income was $88 million or $1 per share versus $1.27 pressure for the second quarter. On a non-GAAP basis, excluding restructuring charges of $0.27 EPS impact, our third quarter net income per share was $1.27, which compares to $1.65 in the second quarter on a comparable basis. The complete reconciliation of GAAP to non-GAAP numbers can be found at the back of the presentation. Turning to Slide 14. I'll review the balance sheet and cash flow summary. Cash and marketable securities were $717 million, down slightly from $740 million at the end of last quarter. Accounts receivable trade balances increased by $324 million quarter-over-quarter due to outstanding invoices related to AVSR and Topaz and higher shipments of third-party module customers. Note, AVSR and Topaz amounts were collected in October. Our unbilled accounts receivables decreased by $37 million due to amounts billed for the Agua Caliente project. Inventory, including balance of systems, decreased $71 million, primarily due to higher module shipments to third parties and increased system project installation of balance of system components. Project assets increased by $91 million as construction activity ramped for our Amherstburg, Belmont, Walpole and Maryland Solar Projects. Deferred project costs increased by $68 million, primarily due to continued ramp of construction at Desert Sunlight, partially offset by initial revenue recognition at Topaz. As a reminder, when a sale of the project -- when we sell a project, the project assets turn either into cost of goods sold or deferred project costs, depending on whether all the applicable revenue recognition criteria have been met. To date, we have not recognized any revenue on Desert Sunlight because all the revenue recognition criteria have not yet been met. However, we are receiving cash payments for this project as milestones are achieved. Our debt increased slightly to $530 million versus $519 million at the end of the prior quarter. Operating cash flow for the quarter was $22 million, and free cash flow was negative $40 million. Year-to-date, we have generated operating cash flow of $435 million. Excluding LPM and restructuring activity, year-to-date operating cash flow is approximately $0.5 billion. We spent $57 million for capital expenditures, a decrease of $100 million from the last quarter. This expected decrease is reflective of continued capital commitments -- completed capital commitments related to previously planned capacity expansion. Depreciation was $66 million compared to $64 million last quarter. This brings me to our updated guidance for 2012, starting with the assumptions behind our guidance as shown on Slide 15. The expected 2012 demand is approximately 1.9 gigawatts DC. We are updating our average module manufacturing cost range to $0.68 to $0.70 from $0.70 to $0.72 per watt in 2012, largely due to improved factory utilization. Our expectation for average module efficiency is between 12.6% and 12.7%. We expect our effective tax rate to be in the range of 15% to 17%, excluding the impact of restructuring and the manufacturing excursion charges which is unchanged from our prior guidance. Now turning to Slide 16. First Solar is updating 2012 guidance as follows: We expect net sales of $3.5 billion to $3.8 billion compared to prior guidance of $3.6 billion to $3.9 billion. As Jim mentioned, we have had several construction sites and corresponding supply chains impacted by the devastation of Hurricane Sandy, and we're assessing the impact and related impact to the projects. In our prior guidance, we had anticipated that these projects would achieve substantial completion and financial close in mid-to-late December. While we have not been able to completely assess the scheduled impact of each project, it is clear that there will be an adverse impact. To accommodate for this timing risk, we have reduced slightly our full year net sales forecast. We would emphasize that this change is merely an issue of timing, which if we're unable to recover on the scheduled impact, we'll move revenue recognition from 2012 into early Q1 2013. We are narrowing the guidance for earnings per fully diluted share to $4.40 to $4.70 compared to prior guidance of $4.20 to $4.70, in each case excluding the charges related to restructuring and manufacturing excursion. The narrowing or the range and consequentially higher midpoint is primarily attributed to our lower full year average manufacturing cost per watt range of $0.68 to $0.70 highlighted in previous slide. The cost improvement stems from better overall plant utilization. Operating cash flows are now expected to be between $650 million and $850 million compared to our prior guidance of $850 million to $950 million. The financial closings and therefore cash receipts associated with the projects risking slipping into the first quarter of 2013 are scheduled to occur after the project achieved substantial completion. We will update our guidance range to accommodate for the potential project delays that may push certain cash receipts into early Q1. Please note as a reminder, operating cash flow guidance includes the impact of cash outlays associated with the announced restructuring activities and manufacturing excursion remediation. Excluding the full year cash outlays associated with these activities, 2012 operating cash flow would be approximately $150 million higher. We are maintaining our 2012 CapEx guidance. As we mentioned earlier in the call, we expect to incur between $20 million and $25 million in restructuring charges in Q4, which will result in full year charges consistent with our April 17 announcement. Now as it relates to guidance relative to 2013, please note that we will announce our full year 2013 guidance in conjunction with our full year 2007 earnings call in February. While we have approximately 90% of our first-half 2013 plant supply already booked, given the timing of projects in our systems business and booking activity for the second half of 2013, we believe we'll be able to provide a higher-quality guidance on our earnings call in February. Now moving to Slide 17, I'd like to summarize the progress this quarter. We continue to execute on our long-term strategy and expand our global presence with the addition of multiple geographies to our portfolio of sustainable markets. Our R&D team is driving fundamental technology investments with the objective of improving our convergence efficiency and cost roadmaps and accelerate their achievement. In addition to module performance enhancements, cost reductions and our vertically integrated offering, we're increasing differentiation of our offering through grid management technology that is difficult to replicate. We continue to have one of the strongest balance sheets in the industry, and our cash position is expected to increase in Q4 as receipts from several projects are realized. Based on year-to-date performance and expected Q4 execution plans, we're increasing the lower end of our 2012 non-GAAP EPS guidance to $4.40. With this, we conclude our prepared remarks and open the call for questions. Operator?

[Operator Instructions] We'll take our first question from Stephen Chin with UBS.

Jim and Mark, just a question on the $8.9 billion backlog. I appreciate you sharing that update. Just curious what percentage of that do you think you can turn in 2013 and what is the sales backlog that you're going to try to plan to manage the business through next year?

Stephen, let's wait until February then we go to that level of detail around guidance. I mean, I think you have to think about what we said in the script was with the first half of next year is firming up pretty nicely relative to our production plan. We just need to get better inside onto next year and the second half of the year, and we'll give you more color around that in February.

We'll take our next question from Sanjay Shrestha with Lazard.

Mark, I just want to clarify. I think you said that you collected about a meaningful spend of cash from AVSR and Topaz in the month of October. Is that correct?

Yes. So that the increase and they are largely was driven by those 2 projects of which the cash came in the month of October.

We'll take our next question from Amir Rozwadowski with Barclays.

Just following up on the prior question on that sort of $8.9 billion in revenue backlog, I was wondering if you could give us any update in terms of what sort of -- broadly sort of the margin structure you guys are seeing? I mean, you have provided sort of longer-term guidance before. And I just wanted to see if there's any variation now that you've sort of built -- sort of what that expectation is around that $8.9 billion?

That -- we have not provided any sort of information or detail on the margin structure of that. And frankly, I'm not sure that's something that we would be willing to do. Suffice it to say, I think everybody understands that with general industry conditions, there is definitely pressure on margins. And new business that we are booking today is clearly not at the same margin levels as historical business. I don't think there is much mystery about that dynamic occurring in the industry. The importance of this information was there was a great sense earlier in the year that overall activity levels have collapsed to de minimis levels. And that simply has not been the case, and we thought it was important to let people know and allow people to understand that we continue to move along at a fairly robust clip in terms of overall activity levels. We'll live with that -- sort of the detail on margin becoming evident as it rolls into recognized revenue and profitability of the business.

We'll take our next question from Brian Lee with Goldman Sachs.

Another one on the expected revenue. Can you talk about the composition of the $9.4 billion that you had entering the year versus the $8.9 billion that you're currently at, at the end of 3Q? I guess I'm wondering how much is modules versus contracted projects versus unsold projects. And are there any non-PPA projects embedded in these numbers?

So the -- if you look at the data as of the end of Q3, $5.3 billion of that is associated with contracted sold projects, okay, systems related business. The other portion is there are some module-associated transactions that we've announced here recently in deals that have come through India comes to mind as a few transactions that we referenced during the call. And then we have included PPA business that is not yet sold, okay. So that's -- so if you look at our normal schedule, we say PPA sold, PPA not sold. It's holistic and it covers both of those pieces, and then it also would include orders that we have recently booked for module-only sales and some of the announcements that we referenced during the call.

And just to follow up on that, one of the dynamics that you will see in these new markets is -- let's take India, for example. A lot of our early activity in India has been module-only sales. That's in part because we're still building our organization and building our on-the-ground capabilities with respect to the systems side of the business. We will also be cautious about how big a project we are willing to undertake in these new markets until we have the level of experience behind this. The best example for that is Australia, where we initially started with a small 10-megawatt project. And once we have built our supply chain relationships and had a clear understanding of labor cost and operating conditions in that market, we will then be able to step up to much larger projects. So as we get our feet on the ground in these new markets, you will see a shift of revenue mix from module only to a more fulsome, more integrated sale over time. So that will impact -- while we're not disclosing in detail, that will impact what the composition of that overall backlog looks like.

And just to make sure, I mean, as you would anticipate, too, by definition, I would describe what is in that $8.9 billion. The vast majority of the revenue value is related to systems.

We'll take our next question from Joseph Osha with Bank of America Merrill Lynch.

Could you help us understand a little bit as you make improvements on the efficiency side, how we should think about balance of systems cost and how that trends through next year?

So what we've said before, I mean, if you look at the relationship between balance of system and the module, think of it as an installed cost perspective, when you have 1/10 of a point improvement efficiency, it's normally going to give you a little bit over $0.01 of benefit to the installed cost.

And the other thing, just to point out and remind people, you -- a lot of people will take the nameplate capacity of the panels, the nameplate conversion efficiency, and do very simple balance of system penalty calculations. And that is a gross and inappropriate oversimplification of how the actual design and performance prediction of the system comes together. So in a very moderate climate, we will have a greater balance of system penalty than we will in a hot desert climate. So you cannot -- you simply can't look at very simple calculations and say, "Here's the balance of system penalty." That penalty -- so-called penalty will be different depending on the exact site conditions of every particular plant.

We'll take our next question from Vishal Shah with Deutsche Bank.

I just wanted to follow up on the revenue expectations you laid out on $1.8 billion of additions [Audio Gap] with strong growth market and what percentages of the market makes?

Vishal, you're breaking up there. Can you please repeat the question?

Okay. I'm just trying to understand what percentage of the $1.8 billion of backlog additions this year came from the U.S. versus from new markets like the Middle East and India? And then you had mentioned earlier, I believe, last year that you expected $3 billion of gross receipts from backlog at that time, which was about 2.9 gigawatts. I'm assuming your revenue composition is similar this time. Can you give us a sense of what kind of gross receipts you expect from the $9.4 billion of backlog?

So on the gross receipts side. The way we highlighted that before was $3.4 billion as it related to anticipated receipt between now and the end of 2014. If you look at what's progressed since that point in time, we gave you that data in April. Really, that's at the end of the first quarter. If you profile that $3.4 billion and if you actually look at the activity that's happened over the last couple of quarters, effectively, that $3.4 billion is all still in front of us between now and the end of 2014. So over the next 9 quarters, you will start to realize that $3.4 billion. Part of it is because this project activity that we've been talking about, a number of the projects we still in our backlog are cash receipts upon substantial completion or financial close, right. So that activity will happen even here in Q4 or Q1. So that will kind of drain some of that working capital that we've got tied up on the balance sheet. But the way -- the best way to think of it, that $3.4 billion is still in front of us and in large will happen over the next 9 quarters. In terms of the bookings and that $1.8 billion revenue-wise, about 1/3 or so of it is actually we'd target as more sustainable markets. It depends on where you classify the U.S., but if you think about the traditional how we're evaluating sustainable markets, about 1/3 is targeted there and other 2/3 is here in North America.

We will take our next question from is Smitti Srethapramote with Morgan Stanley.

Since you presented your target of 3 gigawatts of sales into sustainable markets in 2016 polysilicon pricing and crystalline silicon pricing have continued to collapse faster than I think the most observers had expected. Just wondering if your view on the target has evolved as price of competitive products have fallen.

No, it hasn't. We are very comfortable with our competitiveness at a cost level versus the silicon product. Obviously, their long-term willingness to price at nonprofitable levels impacts the overall market conditions. But one of the things that people need to recognize, a lot of the really distressed silicon pricing is going into the distributed rooftop market. And we simply don't compete in that market. That's a -- there is less brand loyalty. There is less focus on the long-term warranty and long-term power prediction. There is less focus on the bankability in that market, and so that impacts the overall value proposition. Our focus is on a grid-connected utility scale markets and in particular, a big focus on projects that are attracting debt finance and low cost of capital, which impacts sales DOE and we have a very significant market share. I don't have the exact figures, but Bloomberg put some numbers out that showed a fairly dominant market share on our part into debt-financed projects. So it's hard to generalize. We really view the -- a very bifurcated market of the distributed rooftop market largely sold through the retailers and integrators versus the grid-connected market. And so we have not changed our view. We still have the same view of overall PV market size in 2015, 2016, and we still are comfortable with the market share that our business plan projections represents in the context of that overall demand.

And really, what I would say is just that I don't think there's been any change in the trajectory of polysilicon or return to where it's going relative to our estimates when we actually gave our guidance. We believe it's going to be somewhere between $15 and $20. I mean, and it will stay that way for at least the foreseeable future and that's how we think about the competitiveness of our crystalline silicon technology that we go head to head with on a day-to-day basis. So I don't think anything has changed from that perspective.

We'll take our next question from Chris Blansett with JPMorgan. Christopher Blansett - JP Morgan Chase & Co, Research Division: I just want to get a feel for how many megawatts of your 3-gigawatt pipeline you've already recognized revenue on? And then I wanted to ask secondly about the project delays you're experiencing and does this change the subsidy schemes are going to be applied under, willing to go to an ITC versus an ITC cash grant?

In terms of the revenue recognized against the systems business on a megawatt-equivalent basis, I think it's 687 megawatts or something.

626.

626 megawatts on a megawatt-equivalent basis.

And then on the subsidy question, with respect to ITC versus ITC cash grant, I've said this at a number of industry events and in a number of forums. I believe that a broad spectrum of tax benefits are going to get put on the table following the election irrespective of who wins, and I think there will be a robust debate of tax provisions that benefit traditional fossil fuels and tax provisions that benefit renewable including solar. Our position as a company and my position as an individual has long been we want fair and equitable treatment. We want a level playing field. We believe our product needs to make sense economically on fundamentals given a level playing field. Exactly how that debate plays out and what the overall cost structure and tax structure of the industry, including marginal rates, looks like, there is too many moving pieces for me to predict that. But I have an extreme degree of confidence in the overall competitiveness of solar as a part of the generation mix irrespective of where that debate falls out. So it's not something that we view as particularly -- it's an important part of -- that we pay attention to it as part of our business process, but we don't view it as a live-or-die proposition with respect to the industry generally or with respect to First Solar in particular.

We'll take our next question from Satya Kumar with Crédit Suisse. Satya Kumar - Crédit Suisse AG, Research Division: I was wondering just to clarify on the prior question if you can give a sense of how many megawatts. Either AC or DC is fine. So it's your book on a year-to-date basis and your systems business that I can sort of compare apples-to-apples with the 626 number. And how those 2 would sort of look for Q4? I know what you are going keep making revenue in Q4, but I would love to get an update on the expected bookings in the Q4.

So Satya, go back to the first question. You want to understand the number of megawatts related to systems business that we booked for the first half or first 3 quarters of this year? Satya Kumar - Crédit Suisse AG, Research Division: That's right.

Yes, we haven't broken up that detail but I mean, it's not that hard to go back and look at the announcements. We've highlighted them on Slide 6 of the presentation and you can go back and sort of -- or excuse me, not 6, what's slide was that actually?

Slide 5.

Slide 5 of the presentation. That's kind of the listing. Typically, when we announce a project, a systems project, we'll -- it's communicated externally. So I think you have externally available information for the most part to try to at least approximate that, but we have not provided the breakout of the 2.

We'll take our next question from Rob Stone with Cowen and Company.

I wonder if you could just put a little more color on how you're thinking about the margin trend for the fourth quarter since you'll be recognizing a big mix of project stuff and having higher utilization.

In terms of the overall margins for -- on bookings for the quarter, is that was your question is about?

What the influence of making progress on some big projects. You factory utilization. Margins were better than expected this quarter since your cost per watt went down. I'm just wondering how you see the mix and cost utilization trend that's affecting overall margins in the fourth quarter.

So I mean, what we said in the utilization for the fourth quarter is we're going to be running around 90% to 95%, which largely eliminates most of the headwind, right. So if you look at our -- we said $0.64 on a fully utilized basis, but you should use that as kind of a barometer and say, "Okay, we're going to be trending closer to that number because we'll be running for the most part at a -- on a fully utilized basis." So you're going to get $0.01 or $0.02 benefit sequentially with the incremental utilization and then you'll get a little bit of benefit on the efficiency, as we indicated will exit the year around 12.8%. So the -- we'll see a trajectory that it shows that a more improved profile on the cost per watt. We highlighted that on a fully utilized basis and we said, "I think our best plant would be operating around $0.61." So the trajectory continues to move in that direction, and we will take that momentum and we'll move into 2013. And as we indicated in the script that it will be below $0.60 as we trend towards the latter half of 2013.

And our final question will come from Kelly Dougherty with Macquarie.

I just wanted to think about, as you're entering some of these new markets and special ones that have more challenging finance and backstops, are you seeing increased competition from the Chinese -- I mean, do you know what they look like from a balance sheet standpoint? But we see the Chinese saying -- supporting them as they enter some new markets. So I'm just wondering how that plays into the whole competitive mix, especially markets like Chile and things like that that have harder financing?

So again, Jim already alluded to that and I think if you really think about the strength of First Solar, it relates to the bankability and actually was having a conversation with BDN [ph] in Chile just yesterday and she emphasized that point and the importance of that. It's a key differentiator of our ability, the quality of our balance sheet, reliability, ability to spend, our reputation, the same amount of product with our reputation. It's a differentiator in markets like Chile. And when most customers are evaluating -- even though they may have financial backing that is attractive, when you actually look at their financial condition and essentially technically insolvent, knowing that they're going to have to stand behind the performance of an asset for 20 or 25 years in a utility scale type of environment, it's very challenging for them to compete with First Solar given the strength of our balance sheet and reliability and reputation.

And we'll take our final question from Ben Kallo with Robert Baird. Benjamin J. Kallo - Robert W. Baird & Co. Incorporated, Research Division: I just want to get a sense of how many projects were in -- that are involved in the lower end of your operating cash flow, just to see the risk around that?

There is 4 projects. I sort of referenced them in the script when I mentioned that part of the reason our project assets had gone up is because of 3 projects I referenced in Canada and 1 project in Maryland. I mean, those are the projects right now that are subject to potential movement relating to -- once we better assess the impact of Hurricane Sandy. That's not just on the site. It's actually on the supply chain as well because some of our suppliers for those projects are in the impacted area and we've got to make sure that they -- it doesn't -- going to create any disruption to them, their ability to supply us materials that we'll need in order to complete those projects.

And the other thing, just to further clarify, the impact can be in the arena of your available labor force either because direct impacts upon the laborers and their ability to get to work versus their need to take care of their families. Also, in particular, electricians that we may see -- a marshaling of electricians to deal with the massive system damage that's been done and there is -- and additionally, there is -- as power is restored, having been through this in the Gulf Coast, you will discover damage at the building level that will have to be repaired. And so we're just being cautious and prudent, recognizing that we could see impacts that delay us for a number of weeks on the projects that are in that region. We're calling on the same trades that are going to be necessary to restore systems in the impacted area. So it's not so much. I mean, there is some direct impact on our ability to access the site, water on the site, issues such as that. But it's also a broader impact upon the entire infrastructure, including the availability of labor, including the ability of our vendors to perform or the need for our vendors to redirect resources to critical restoration activity. So it's just we're in the middle of an area that's been impacted, and we're trying to reflect caution until we better understand how that's going to flow through to us. As Mark said, we believe it's -- this will largely play out as timing impacts to us, not as ultimate value impact to us.

And the one thing, just as I referenced in the script, that essentially, those projects have to reach substantial completion in order for a financial close to happen. Substantial completion also includes connection to the grid, right? And so new connections are not going to be prioritized over the issues they've got to restore existing power, right? So we're already starting to see where we initially had scheduled activities. Things are starting to slip. We don't have a commitment right now around the ability to get a project in particular connected as we initially thought we did because it's the uncertainty of what's going on right now.

And that does conclude today's conference. We thank you for your participation.