TPI Composites Inc (TPIC) 2017 Q2 法說會逐字稿

Good afternoon, and welcome to TPI Composites Second Quarter 2017 Earnings Conference call. Today's call is being recorded, and we have allocated an hour for prepared remarks and Q&A.

At this time, I would like to introduce your host, [Anthony Rozmus], Investor Relations for TPI Composites. Thank you. Sir, please go ahead.

Thank you, operator. I'd like to welcome everyone to TPI Composites Second Quarter 2017 Earnings Call. In addition to our press release, you can find our Q2 earnings slide presentation on our IR website.

Before we begin, let me remind everyone that during this call, TPI Composites' management may make certain statements that constitute forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. These include remarks about future expectations, projections, beliefs, estimates, plans and prospects. Such statements are subject to a variety of risks, uncertainties and other factors that could cause actual results to differ materially from those indicated or implied by such statements. Such risks and other factors are described in our Form 10-K and other periodic reports as filed with the Securities and Exchange Commission. The company does not undertake any duty to update such forward-looking statements.

Additionally, during today's call, the company will discuss non-GAAP measures, which we believe can be useful in evaluating our performance. The presentation of this additional information should not be considered in isolation or as a substitute for results prepared in accordance with GAAP. The reconciliations of GAAP to non-GAAP information can all be found in our earnings release, which is posted on our outside, at www.tpicomposites.com and is also included in our Form 10-Q as filed.

With that, let me turn the call over to Steve Lockard, TPI Composites' President and CEO.

Good afternoon, everyone, and thank you for joining our second quarter 2017 earnings call. I'm joined today by Bill Siwek, our CFO. I will start with some highlights from the second quarter, followed by a brief update on the wind market and TPI's progress on our strategy of strong and diversified global growth. I will then turn the call over to Bill to review more details on our financial results. I will then conclude with a review of our 2017 outlook before we open up the call for Q&A.

Please turn to Slide 5. TPI delivered another strong quarter, posting record financial results, which also exceeded our plan for the 3 months ended June 30, 2017. Our net sales experienced healthy year-over-year growth, increasing 27.8% to $248.2 million, while total billings increased 17.8% to $231.1 million. Adjusted EBITDA for the quarter increased 47.9% to $30.8 million, and adjusted EBITDA margins expanded 170 basis points to 12.4%. The strength in adjusted EBITDA was driven not only by an overall increase in our volumes, but as importantly, by our relentless efforts toward continuous operational improvements and driving down costs across the board. The estimated gigawatts generated by the Q2 '17 sets we delivered once installed is estimated to be 1.6 gigawatts compared to 1.3 gigawatts in Q2 '16.

Our financial performance over the last several years speaks volumes to the effectiveness and dedication of our employees, our deeply collaborative business model and the execution of our growth strategy. As we discussed in our Q1 call in April, we announced a multi-year supply agreement with Vestas to produce wind blades at a new facility in Matamoros, Mexico, which we expect will reliably and cost-effectively serve the rapidly growing Latin America market, including Mexico, Central and South America, by truck and rail as well as by water from the Port of Brownsville, Texas. Today, we announced the signing of a multi-year supply agreement to provide wind blades from 2 manufacturing lines at TPI's facility in Taicang, China, with production expected to commence in the first quarter of 2018. This agreement will allow us to backfill capacity in that region that will free up at the end of the year.

I'd also like to note that we were pleased to have completed our secondary public offering of 5.1 million shares in May, which has helped us improve liquidity of our shares in the public marketplace and expand our investor base.

Finally, we announced yesterday the appointment of Joe Kishkill as our Chief Commercial Officer effective August 21. Joe will lead our diversified growth strategy, including all commercial aspects of the wind blade business and TPI's expansion into new markets for advanced composite structures. We are excited to have Joe join our executive team. His international business development and executive experience in both conventional and renewable energy resources provides great tools to help us to continue to grow and diversify both our wind business and new applications for advanced composite solutions around the globe.

This is an exciting time at TPI. We are pleased to have signed new 2 new supply agreements as well as demonstrating our continued conversion of our robust demand pipeline. We are confident in reaffirming our target of a 25% revenue CAGR through 2019. We will continue to focus on diversifying our sources of revenue across customers and geographies, and we'll continue to take advantage of the growth in the global wind market, long-term stability in the U.S. wind market and the ongoing wind blade outsourcing trend.

If you turn to Slide 6. As of today, we have 48 manufacturing lines of the long-term supply agreements, providing long-term revenue visibility of approximately $4.4 billion through 2023. This includes the 2 new manufacturing lines just announced and the 7 lines for GE that will expire at year-end. While the minimum guaranteed volume under contract is approximately $2.8 billion, our contract structure encourages customers to purchase the maximum volume as their price increases if their volume drops below the maximum. Our pipeline also remains strong as we have prioritized 28 additional molds to close over the next 24 months. As blades increase in length in order to continue to drive down levelized cost of energy, TPI's revenue per year per line is increasing. Therefore, we do not need to grow the number of lines by 25% in order to achieve our 25% revenue CAGR target through 2019.

As shown on Slide 7, annual installed onshore wind growth is expected to increase from 50.1 gigawatts in 2016 to 62 gigawatts in 2026. This growth will be driven primarily by developing markets, which, according to data provided by MAKE, will grow at a CAGR of 8.8% during that period. While more mature markets, those with at least 6 gigawatts of installed capacity at the end of 2016, will continue to grow but at a more modest CAGR of 0.4%. During that same period, MAKE forecasts that the annual share of overall installed capacity from developing markets will increase from 15.2% in 2016 to over 28% in 2026. TPI is well positioned to serve these developing markets from our facilities in China; Juárez; and Matamoros, Mexico; and Turkey. And we expect to growth of these markets will continue to drive the outsourcing trend we've seen over the last 10 years.

With respect to the U.S. onshore market, as you can see on Slide 8, the next 4 years are expected to be the most stable in the history of the U.S. wind industry with expected annual installations averaging 9.7 gigawatts. According to UWEA, across the U.S., 29 wind projects representing a combined 3.8 gigawatts announced that they either began construction or entered advanced development in April through June 2017 for a total of 25.8 gigawatts of wind projects currently underway. That is up more than 7.5 gigawatts from the 18.2 gigawatts underway as of 1 year ago. Also, in the second quarter, corporate customers represented 37% of total project capacity contracted with 6 major commercial customers, including Akamai Technologies, Apple, General Mills, Goldman Sachs, Partners HealthCare and T-Mobile, all signing U.S. wind PPAs for the first time. Additionally, 2 offshore wind projects in Maryland were awarded offshore renewable energy credits, a key step in the development of offshore wind in the U.S. We believe we are very well positioned in the U.S. with our current customers accounting for 99.8% of the U.S. market share in 2016 and an expectation that these customers will continue to garner a significant amount of U.S. market share in the future.

Turning to Slide 9. Beyond 2020, we still see a very strong market for wind energy both in the U.S. and globally driven by a number of factors including: Overall competitiveness of wind energy. LCOE is at or near grid parity with natural gas in many regions on an on-subsidized basis, and the industry is continuing to drive LCOE down through innovation and scale.

Commercial and industrial demand. Currently 102 RE100 companies have made a commitment to purchase 100% of their electricity from renewables resources including major corporations, such as General Motors, Coca-Cola, BMW, and a recent addition to the list, JPMorgan.

Utilities have identified large-scale renewables as a key growth area, and the rapid growth of renewables is spurring investments in transmission, other grid technologies and cost-effective energy storage that will ease the integration and expand our asset base. Recent examples include the Wind Catcher Energy Connection that was recently announced by American Electric Power. That will include a 2-gigawatt wind farm in Western Oklahoma with 800 turbines and a dedicated 350-mile single circuit 760-kV power line from the wind farm to enable AEP to provide for power to customers in the states of Oklahoma, Arkansas, Louisiana and Texas. This is on the heels of Berkshire Hathaway Energy announcing 4 gigawatts of wind projects during the last year in Iowa and several Rocky Mountain states.

Offshore. The recently recorded price trends in offshore wind have indeed led to a dynamic revolution in the global landscape for offshore wind energy. The increasing urbanization and economic development, which have subsequently boosted electricity demand, is also playing a major role in fueling offshore wind market trends. According to MAKE, the offshore market is projected to grow at 17% CAGR through 2026.

Repowering will become a larger share of onshore opportunities. In the U.S., MAKE estimates 7 to 8 gigawatts through 2020 and also estimates that 28% of global repowering will occur in the next 5 years, with Germany, France, The Netherlands and the U.S. accounting for 83% of that volume.

Decarbonization. The reduction of greenhouse gas emissions from the electricity sector as a response to climate change is driving decarbonization globally. This includes the rapid expansion of variable renewable energy solutions for intermittency and the evolution of power market design.

Vehicle electrification. Electric vehicles represent less than 1% all-new automotive sales in 2016. But EV sales are increasing, and over 100 new EV models are poised to hit the market by the end of the decade. Meanwhile, battery prices are declining fast enough to put EV purchase prices on par with traditional internal combustion vehicles during the 2020s, and increasingly prevalent charging infrastructure will reduce, and ultimately, eliminate ranging anxiety as a barrier to adoption. EVs could serve as one of the largest sources of new electricity demand, especially in Western markets that have seen minimal load growth for decades.

Energy access. As many as 1.2 billion people remain without access to electricity, severely handicapping their economic and social development, more than half of the off-grid population live in Sub-Saharan Africa, which, despite strong progress elsewhere, retains an astoundingly low 45% electrification rate. And much of the population in emerging countries with access to electricity still suffers from frequent unpredictable outages. Providing affordable, reliable energy to this population is widely recognized to be among the most formidable and most important development challenges today. And the pace and manner in which energy access is provided to more of the population will have dramatic impact on power demand, generation sources and greenhouse gas emissions.

As we stated on our prior earnings calls, and you've likely heard from our customers, we believe the current market conditions that are impacting our customers, including the general trend in many markets of transitioning to auction-based systems, U.S. market demand shifts driven by the current PTC cycle and increasing competition from solar in many regions, are putting pressure on the industry and our customers to continue to drive down LCOE. These market dynamics will result in more product transitions in late 2017 and 2018 than initially anticipated, and therefore, likely reduce our year-over-year revenue growth rate in 2018. But expect, it will lay a strong foundation for 2019, enabling us to deliver on our target of 25% revenue CAGR through 2019.

Let me now turn the call over to Bill to discuss our financials in more detail.

Thanks, Steve.

If you turn to Slide 11. Net sales for the quarter increased by $53.9 million or 27.8% to $248.2 million compared to $194.3 million in the same period of 2016. Net sales of wind blades increased by 31.1% to $239.8 million for the second quarter of 2017 as compared to $182.9 million in the second quarter of 2016. The increase was primarily driven by a 36% increase in the number of wind blades delivered during the second quarter compared to the same period in 2016, primarily from our China and Mexico plants, partially offset by a decline in the average sales price of the same blade models delivered in both periods as a result of geographic mix and savings in raw material costs, a portion of which we share with our customers, as well as foreign currency fluctuations in China and Turkey.

Total billings for the 3 months ended June 30, 2017 increased by $34.9 million or 17.8% to $231.1 million compared to $196.1 million in the same period of 2016, driven by a 26% increase in wind blades manufactured. The impact of the currency movements on consolidated net sales in total billings were reductions of 2% and 1.9%, respectively, for the quarter.

Gross profit for the quarter totaled $34.6 million, an increase of $11.7 million over the same period of 2016, and our gross profit margin increased by 220 basis points to 13.9%, notwithstanding the fact that we had 9 manufacturing lines in startup during Q2 of '17 compared to just 3 lines in transition in Q2 of '16. The increase in gross margin was driven by lower cost of goods sold as a result of improved operating efficiencies, the impact of savings in raw materials and foreign currency fluctuations.

Startup and transition costs in the second quarter were $10.5 million as compared to $3.1 million during the same period a year ago. The increase in costs for Q2 2017 are related to our new plants in Mexico and Turkey and the startup of a new wind blade model for one of our customers in China. Before the impact of the startup and transition costs, our gross profit margin in the quarter was 18.2% compared to 13.3% in the second quarter of 2016, reflecting continued margin expansion from our fully operational facilities.

General and administrative expenses for the quarter increased to 4.3% of net sales or $10.8 million as compared to 2.7% or $5.3 million in 2016. The increase was primarily driven by share-based compensation of $1.7 million recorded in the 2017 period, the cost of our secondary offering as well as additional costs incurred to enhance our corporate support functions to support our growth and public company governance.

Net income attributable to common shareholders was $13.9 million or $0.41 per share on a fully diluted basis compared to $9.1 million or $2.15 per share on a fully diluted basis in the same period a year ago. The increase in net income is a result of the improved operating results described above.

Adjusted EBITDA was $30.8 million compared to $20.8 million in Q2 of 2016. Our adjusted EBITDA margin for this quarter was 12.4%, a 170 basis point improvement from our margin of 10.7% in the second quarter of '16, again, notwithstanding $7.5 million more of startup and transition costs in Q2 '17 compared to Q2 '16. Before the impact of startup and transition costs, our adjusted EBITDA margin was 16.6% in the second quarter of 2017 as compared to 12.3% in the second quarter of 2016.

We invoiced 692 wind blade sets in the quarter compared to 551 sets in Q2 '16, and the estimated gigawatts to be generated by the Q2 '17 sets once installed is estimated to be 1.6 gigawatts compared to 1.3 gigawatts in Q2 '16. At the end of the quarter, we had 46 manufacturing lines dedicated under our long-term supply agreements with 39 lines installed, of which 30 were in full operation, and the remaining 9 lines were in various stages of startup. The dedicated lines that we announced today are expected to be installed beginning in early 2018 and bring our total dedicated manufacturing lines to 48 as of today.

Turning to Slide 13. Cash and cash equivalents as of June 30, 2017 were $130.8 million, and total indebtedness was $130.4 million, resulting in net cash at June 30 of approximately $467,000 compared to a net debt of $6.4 million at year-end. For the quarter, we generated cash from operating activities of $18 million while spending $9.8 million on CapEx, resulting in free cash flow for the quarter of $8.2 million. Consistent with our past practice, we'll continue to finance a relatively modest portion of our CapEx for our new facilities, and the balance will be funded with cash flow from operations.

We continue to be pleased with the strength of our balance sheet and our ability to continue to generate the cash we need to expand our global footprint, and we believe we are in a solid position to continue to capitalize on the growth opportunities before us.

I'll now turn the call back over to Steve.

Thanks, Bill.

Please turn to Slide 15. Now I'd like to give an update on our guidance for the balance of 2017. We're going to provide some additional quarterly guidance to help you all to better understand some of the nuances of our production cycles and the slight seasonality in our business based on holiday periods.

We are reaffirming our guidance for total billings for 2017 of between $930 million and $950 million. With respect to sets, we will still be in the range of our guidance and expect to deliver between 760 and 770 sets in Q3 and between 700 and 710 sets in Q4. ASP will be within the original guidance range but trending toward the top of that range during Q3 and into Q4.

Dedicated manufacturing lines at year-end will be below the range for a couple of reasons. First, at the time we provided this guidance, we did not anticipate that GE would choose to not extend our agreement with them for 4 lines in China. Second, as the market dynamics, such as auctions discussed earlier, began to manifest, our customers and potential customers have been looking at somewhat smaller deals, for example, 2 lines with options for additional lines versus committing to 4 or 6 lines immediately. Notwithstanding fewer lines being committed up front as blades continue to get longer, the ASP on those blades continues to go up. Furthermore, as we continue to drive operational excellence through our plants, we are able to keep throughput for the larger blades at or above historical throughput levels. So our revenue per line will increase from approximately $25 million per year to approximately $35 million per year when at full run rate. Therefore, the number of new lines needed to reach our target revenue CAGR of 25% through 2019 is lower than in the past. So although transitions and startups impact us in the short term, long-term growth and profitability is ultimately enhanced, and the flexibility we are able to demonstrate to our customers strengthens our relationships for the future.

Total lines installed at year-end will be 37 versus our guidance of 40, driven primarily by the 4 lines per China that will be removed at or near year-end.

Given the discussion earlier about transitions and startups, we anticipate startup and transition costs to be at the high end of the range or slightly above for the full year. We expect these costs to be between $13.5 million and $14 million in Q3 and between $11 million and $11.5 million in Q4.

Capital expenditures, effective tax rate, depreciation, amortization and interest expense should all be within the guidance ranges. Income tax expense is expected to be higher, driven by higher forecastable taxable income. Share-based compensation will be below the range as no equity no grants other than 2 new employees have been given -- have been made this year.

We remain very confident in our global competitive position and the application of our dedicated supplier model to take full advantage of the strength in the growing regions in the wind market, the trend toward blade outsourcing and the opportunities for market share gains provided by the current competitive dynamic.

It was another successful quarter for TPI.

To recap, we delivered outstanding results both on the top line and on an adjusted EBITDA basis; signed a new multi-year supply agreement in China to backfill the expiring GE contract; expanded our relationship with Vestas with a new multi-year supply agreement for a new manufacturing facility at Matamoros, Mexico to serve the Latin American market; completed a secondary offering; and hired Joe Kishkill into the newly-created role of Chief Commercial Officer.

We are very excited for what's to come, and I want to thank the TPI employees for their hard work and our shareholders for their continued support.

As a reminder, we'll be hosting our first Investor Analyst Day on Friday, November 17 at the Roosevelt Hotel in New York City. Formal invitations will be sent out in the near future. But if you have any questions or interest in attending, please don't hesitate to reach out to our IR team directly at investors@tpicomposites.com.

Thank you, again, for your time today. And with that, operator, please open the line for questions.

(Operator Instructions) Our first question comes from the line of Paul Coster with JPMorgan.

So in China, is it a new customer or an existing customer?

Yes, Paul, it's Steve. Paul, we're, at this point, not able to talk about the specific customer situation around the new contract for China. As you know, our target is the top 10 global players. You can imagine the customer would be on that list. But we're not going to get out ahead of them in terms of announcement and timing. And so at this point, what we can say is more information will follow shortly. But I think given the timing of this call, we wanted to make sure folks understood we have reached an agreement to backfill. This is an important step for us, as you know, to backfill the lines that'll be coming out of China at the end of the year, and I think that's what we can say at this point.

Is it a like-for-like swap in terms of the configuration of the blades? Or is it a larger blade? And -- you still have a couple of more lines there, don't you? So does it backfill on revenue until the [yield] opportunity for additional business there? Or just take us through that please.

No, thanks, Paul. It's a good question. So as blades get larger, as we move into, call it, 65 or 70 meter-class blades whereas some of the lines coming out and being transitioned within TPI are more in the older 50-meter class, if you will, the revenue per year per line is going up by quite a bit. And so one way to think about this -- as you know, Paul, we're not going to get into the specifics of one customer or one factory so much. But I think, generally, you could view it as something like what used to be 7 molds could be replaced by something like 4 lines. And that's in the case of, for example, when the blades are very large, if they're using more advanced materials, where the bill of material cost maybe a bit higher, that's an example of what can happen. So perhaps, in the 4 to 5, mold lines would replace what would have been 7, generically, as a way to think about it. Hope that's helpful.

Yes, it is. But just a confirm. You still have 2 lines available in China even with this backfill, correct?

No, I think the existing facility that used to be a 4-mold facility, we would not put 4 larger blades in. We'd probably put in 3, would be a way to think about it. So that facility probably has room for, one more. But in that case, 3 lines replacing what was the space for 4, in that case.

Got it. And you still have good prospects sort of backfilling on that final line as well?

We do. That -- we converted 2 additional lines out of the 28-mold backlog or pipeline, as we described in our last call, and added 2 more back in. So we're still at a pipeline of 28 molds and plenty of demand in Asia to continue to work on both backfill, if needed, and growth.

Okay. My last question. You told us that repowering and with closing more product transitions. So I think what we are inferring there is that once again it's the next -- the existing turbines have been larger blades installed on the more different configuration, and so that means the molds can be swapped out. Can you quantify the impact in '18 for us in terms of the transit number of lines that are likely to go through this transition? And then just cap it by, remind me as what the -- in terms of this 29% (sic) [25%] revenue CAGR through '19, what's the expected adjusted EBITDA CAGR through '19?

Yes, Paul, thanks. So a couple of pieces there. So I think on the repowering question, repowering is helping to drive some of the gigawatts that are being installed or planned to be installed over the next few years both in the U.S. but then also outside the U.S. And in some cases, you may remember, Paul, some of that is replaced with, perhaps, only a portion of the machine. Maybe just the blades are being swapped out, in which case, we may go backwards and build some older models, smaller blades than our standard. And in other cases, the entire machines are being upgraded to brand-new machines, which to us looks just like new business. So we think of repowering a bit more as a market driver that helps to grow the gigawatts installed on an annual basis. And it's one important -- it was on our list of market drivers. It's one important driver that helps contribute to that. What I'd say about the product transitions just generally is there are couple of things that are driving that. One, right now, as you think about in the U.S. market the PTC cycle, there is a need -- and for auctions around the world as well as in the U.S., there is a need for wind to continue to be more and more competitive, and there is a competitive tension between our customers as they respond to those market conditions. So -- and you'll remember, larger blades and taller towers is one of the main drivers to reducing LCOE for wind turbine. So our product transitions, I wouldn't say they're being driven by repowering. They're really being driven by this quest to continue to drive down levelized cost of energy. Because of the timing, there are more transitions that are being queued up late this year and into next year than I'd -- I think would be normal. It's, call it, just an abnormally high number of transitions. So there'll be some short-term impact on revenue and profitability as we invest in those transitions. The really good news is we're driving for even stronger demand in '19 and '20, more competitiveness on behalf of our customers. And then as the PTC phases out and we get into 2021 and beyond, wind is going to be that much more competitive. So we see it as a smart investment on behalf of our customers and ourselves to carry that out. And then, Bill, do you want to comment any more specifically about a more specific part of Paul's question?

Yes. Paul, on the adjusted EBITDA CAGR, we haven't provided that in the past and don't plan on doing it today. We may -- we will likely touch on that at our Analyst Day in November.

Our next question comes from the line of Stephen Byrd with Morgan Stanley.

Wanted to hit on just a couple of things. Paul hit on a couple of key things. But the -- in terms of just at a high level -- your customer usage of the facilities in terms of -- you've mentioned many times the natural incentives for them to go to the maximum output level onto the contract. Are you generally seeing any trends in terms of the utilization among customers? Are they sort of hitting that incentive or are there reasons why they might not do so?

Yes, Stephen, we're generally operating still on the high side of that range, if not at or even above at times, of the plant capacity on a global basis. And the general reason when we're not, if we're not, is because of a transition. So that's generally still the case. That's been the case. And as we've grown so dramatically over the last number of years, as you can imagine, most all the factories have been basically sold out, kind of ramping and chasing that volume, if you will, and remaining on the high side of the plant capacity. That's still pretty much the case. What is changing a little bit here, again, is the -- are the accelerated transitions or the number of transitions that are happening at a similar time, not -- generally not otherwise. From time-to-time, there'll be a bit of a dip in volume for one customer in 1 quarter. But generally speaking, we're operating on the high side, except for transitions.

Understood. I wanted to check on South America as well. That seems like an important opportunity in the long run. And seems like Mexico has served you well, but I did want to check. As the South American market gets bigger and bigger, how do you feel about locations in Mexico relative to being closer than South America? I think there have been issues in the past where thinking about being directly located in places like Brazil, but I did just want to check that we should expect that Mexico could grow over time as the South American market grows.

Yes, our plan for Matamoros is to serve the South American market. Mexico, we described is the Latin American market, right, Mexico, Central and South America. But think of it, Stephen, it's kind of ex Brazil. It's probably the simple way -- simplest way to think about it. Brazil has local content requirements. It's a big market at times. But it served generally locally from Brazil for Brazil. So for us, to build blades in Matamoros and work through the Port of Brownsville, we can touch the Southern Mexico, Central and South America very cost-effectively by water. And so you can imagine by truck and rail, we would serve the northern part of Mexico, and then by water, serving the rest of, let's call it, the Latin American market on primarily ex Brazil, is the way to think about it. And yes, it's a growing and emerging region. It was the next significant foothold that we chose to build out, and we're pleased to have Vestas as our anchor customer there. We do see it is an important growth region for us and another cost-effective hub to grow from in Matamoros.

Great. And then just lastly. Margins were very good, and one of the elements you mentioned was materials costs. Are there any particular trends with respect to materials costs? Is that a longer-term dynamic we should expect? Or is it a more short-term benefit that you realized?

Yes, Stephen, I wouldn't -- I would certainly not call it a short term. I mean, we are -- it's something we focus on every day. Because again, this is one of the key drivers for us driving down LCOE as an industry. So I see it as a longer-term trend. Early indications from some of our work for this year would suggest that next year will be another good year for us from that perspective. And again, as we gain scale, the volumes certainly help. And I think some of the consolidation in the marketplace will help us as well as we become a larger and larger player. So I think all of those factors are helping to drive the continued trend that we've seen on a downward motion.

Ladies and gentlemen, our next question comes from the line of Phil Shen with Roth Capital Partners.

I want to just start with -- I think, Steve, in your prepared remarks, you talked about product transitions through the rest of '17 impacting '18. It seems like the U.S. market is a bit soft, given some of the commentary from other players in the wind ecosystem. And you mentioned that it might impact 2018 year-on-year revenue growth. So I know you haven't provided official guidance for '18, but was wondering if you might be able to quantify the range of the potential impacts.

Yes, Phil, fair question. It's a bit early, I think, for us to do that. We'll plan to provide some more clarity on targets in the Analyst Day later in the year. I think the general way to think about it is we do expect to grow in '18. And the growth in product transitions, number of lines under contract, the number of lines will be bringing up, along with the higher revenue per year per line, provides a really strong basis for the 2019, 25% through 2019 CAGR. So we'll grow in '18, but we don't think that growth will be at 25% year-over-year. So we're just trying to point out that the '18 number will be less than 25% year-over-year, but the investment that's being made there strongly supports the 25% target CAGR through 2019.

That detail is helpful, Steve. Again, on the U.S. softness. Can you just help us understand how you guys are able to weather the softness in Q2, Q3 and due in PTC cycle when some of the other -- your other peers downstream, if you will, are experiencing some challenges there. I think I know the answer to that question. It may have been related to Stephen's question earlier in terms of being able to be at high utilizations, so forth. But wanted to kind of ask that question directly because I know that's a topic on some investors' minds.

Yes. Well, we're certainly a global company, and each of our operations serve as large and cost-effective an addressable market as we can set them up to serve. The Iowa factory is pretty much focused on the U.S. market as you would imagine. But all of our other plants around the world serve markets that are far beyond just the U.S. market, including our Mexico site for that matter. So I think just the first point is our dedicated supplier model kind of overall volumes that are contracted for and volume-based pricing in cost-effective hubs gives us a base to work from. Again, we become a bit more of a baseload supplier in that sense, and some of the swing capacity might be elsewhere at times. But again, we've been working hard to diversify the sources of revenue geographically. I think that's paid dividends. And wind markets -- any individual market is up or down in a short term, we've designed it to have less impact on our company, and I think you're seeing those results. Now those can also be -- those soft periods can also be times to work through transitions, and that is a little bit of what you're seeing. So we're utilizing with our customers a bit of that timing softness, if you will, to make the investments in transitions. So you are seeing some of that as well.

Great. One last one. I'll pass it on. In terms of the Siemens Gamesa merger, as they look to rationalize their product portfolio and optimize their supply chain post-merger, can you talk about your ability to maintain your relationship with them? And perhaps just -- perhaps give us an update or some insight into how that -- how those discussions might be developing.

Yes. Thanks, Phil. So as we've answered before, we're not going to get out ahead of our customers in terms of giving planning details or information they might view as being sensitive or confidential. I think suffice to say at this point what's public is that we have a strong relationship with the Gamesa side of the house. Ricardo Chocarro, a Gamesa person, has been named to lead the land base business on a global scale. We know Ricardo well, and he and his team know TPI well. We also have an agreement out of Turkey for one line with an option for one more line. And Siemens Gamesa was awarded the Turkey tender last week as well. So you can imagine those dynamics would bode well for our future. We still see an opportunity for more outsourcing in general across the industry. The GE-LM acquisition has been a bit of an exception to that trend, as you know. But generally, we're seeing more opportunities for those types of moves, and we would expect and hope that we can earn more business out of the combined Siemens Gamesa team going forward.

(Operator Instructions) Our next question comes from the line of John Quealy with Canaccord.

First question on China. I'll try to go about it at a different way to the new or existing customer question, just maybe broader. How are those conversations going with industry participants, LM customers, et cetera? Is the conversation progressing as planned? Or have you touched everybody you wanted to? Just broad strokes. I know you won't give us details, but broad strokes, how is that playing out for you folks?

Yes. I think our dialogue continues as you might imagine. Our target was the top 10 players before the GE-LM acquisition, and our target continues to be the top 10 players, John. And we all know one another. It's a pretty small group. And so yes, we continue to develop those relationships and nurture them. And you can imagine, as we've said before, the 28-mold pipeline that we have is a nice combination of existing customers and growth as well as new customers. We'll continue to chip away and convert off of that 28-mold pipeline.

Okay. Great. Second, maybe for Bill, just on FX in the weaker dollar here. I know it was sort of a -- sounds like a rounding error for the quarter in the 6 months. But talk to us, if you would, about what your impact is expected for the back half of the year. And remind us again on the hedging policy that you folks have, if any.

Yes. Sure, John. As far as the hedging policy, again, we don't have a formal hedging policy in place where we're buying forwards or futures. But we look at natural hedges based on how we're buying raw materials, what our contracts are denominated in and how we manage our debt at the local levels to manage FX risk from that perspective. From an overall impact on operations, it is actually a bit of a benefit over the last 6 months. We've -- it's been helpful on the cost of sales side in China and Turkey, specifically. It's a little bit hard. I don't have my crystal balls, maybe not as clear as everybody else's on where rates are going to go on the back half of the year. But I would suspect that we will continue to see some benefit from those as we move forward.

Got you. Okay. And then 2 other questions for me. First on the bus relationships. I know this is outyear, but can you give us any updates on ancillary markets and in movements in the transportation sector broadly, I would say?

Yes, John, we're continuing to make progress against the development programs that we've discussed in the past and against our production customers in the electric bus and general bus industry as well, continuing to make steady progress there, but have not made any additional announcements at this point.

Okay. That's fine. And then my last question, the Chief Commercial Officer role. Talk about why the timing now for that role, and maybe some of the soft targets or hard targets for the next 12 to 24 months. It might be a little bit more of an Analyst Day thing, but if you could just preview what to expect from maybe some organizational things at TPI.

Yes, thanks, John. So you'll definitely meet Joe at the Analyst Day, and we're really thrilled that Joe has agreed to join us. As with a number of the executive hires we've made in the last couple of years, we're bringing on athletes, pros that have done what we're still aspiring to do and are going to do as we continue to grow our business. Joe ran a $3.1 billion revenue stream one lane over in the solar industry for 3 years for First Solar. He has tremendous international experience as we'd certainly become more and more international in our scope. So we're just adding really solid athletes to each of the key positions. And, John, I've done and led a fair amount of that executive market development work myself in the past as we get bigger and becoming a public company with the various roles we all have as we evolve. It was very important for us at this point in time to add another senior exec to the team. So that's the reason for the timing. We'll continue to build out our team both in terms of the wind market development function, program management as well as the diversified markets piece of this. It's very important for us to continue to grow and diversify, as you've heard many times. So Joe will continue to help us build out that team.

Our next question comes from the line of Jeff Osborne with Cowen and Company.

Just 2 on my end. Most have been asked. Can you -- there is a lot detail about 2018 and, I guess, the growth there. But can you just remind us of kind of the cadence of what's going on operationally, recognizing that you're not going to be giving quantitative guidance, maybe not now but maybe at the Analyst Day. But for example, the lines in Turkey that Gamesa took down and mystery customer in China. My sense is that those are kind of dead in the first half of the year; and second half, loaded. What are -- operationally, what do we think about Matamoros in the first half of '18 vis-à-vis the second half and Mexico 2 and Turkey 2 expansion? Will those be running at full speed in the first half? I'm just trying to think about modeling the second half of '18 versus the first half.

You bet, Jeff. I'll hit -- there are a couple of things in there. First of all, you mentioned Gamesa in Turkey. That mold will get dropped into the facility at the end of this year fourth quarter time frame, so you can envision ramp-up for the first couple of quarters and then taken up to kind of a full ramp right in the back half of '18. We have a couple of new Nordex molds going in, in Turkey. We are actually in the process of installing that first mold right now. The second one will go in towards the end of the year. And so we'll be -- we'll have production volume off of those lines in the first quarter. But again, it won't be ramped up to kind of full-ramp speed until probably the end of the second quarter as well. The new lines that we announced for China, as we mentioned in the call, we plan to begin production sometime in the first quarter, likely towards the end of the first quarter. So you would expect to see kind of full production out of those 2 lines in the back half of the year, probably end of third quarter, beginning of fourth quarter. And then as far as Matamoros goes, as you know, we're building a new facility there. The plan right now would be to have molds in place towards the middle of the year, next year, and then ramping those molds up through the balance of '18. So you'll really see a full production volume out of Matamoros not until early '19.

Got it. And then for -- I know you don't like talking about specific customers, so I'm not sure if you can answer this. But for the extension that GE gave you through 2020, was there any indication of a tip change or a redesign of the mold for the sub-3 years what's left of the contract? Or do you anticipate, as far as you know, it should be producing the same product that they're using today?

Yes, Jeff, we can't get into the detail on one customer and their specific transitions at this point. If customers want to make announcements around their models, they'll do so, but it's really not our place to do that.

Understand. Just broadly, can you talk about cycle times, either one of you? Certainly, China was impressive when I visited there. But can you just talk about some of the other facilities of being able to port the cycle times to the goal of getting a blade made in 24 hours?

Yes, Jeff, I think that we have made great progress over the last 18 to 24 months. Steve mentioned the athletes that we've brought in that have contributed significantly to that effort and in driving operational excellence through all of our plants. I would suggest that with the exception of the new lines that we're ramping, for the most part, we are at or near for that 24-hour goal. There are times when we're below. There are times when we're a little above, and there are some times that we slow down for various reasons. But I think we've made great progress in that area. We'll continue to drive that. And again, that -- part of Steve's prepared comments earlier about being able to drive throughput on the larger molds at the same levels as the smaller molds, that's a perfect example of us driving that cycle time and really driving costs. Driving cycle time also drives costs out of the operation, and that's some of the results of the margin that you're seeing in the last couple of quarters.

Jeff, just to put a little finer point on that. We mentioned our average run rate, $25 million per year per line as being a little bit of a historical number for us in the future, being more like $35 million per year per line. Of course, that's because of the larger blade. But to Bill's point, it's a combination of cycle time and higher -- and larger blade at the same time. And we'll actually have some models running north of the $35 million. Just as in the past, we had some models running a little less than the $25 million. But just to give you average sense of it, if you will, there are some models that were south of $25 million, and other models now that'll be north of $35 million, and that's what helping us to achieve. I mean, it's really an adjustment if you think about as opposed to a reduction in guidance. Think about these number of lines as an adjustment to larger, much larger blades, this idea of 4 lines replacing 7. I mean, that's a pretty significant shift in the way we think about the number of lines required to grow our business is shifting as well. So hopefully, that point came through as being really kind of a fundamental shift to larger blades. It's driven in large part by this quest for LCOE reduction, and to Bill's point, our cycle time management that's helping us get there. Powerful progress.

I appreciate the details. Certainly accommodating your customers' innovation cycles in having to implement those. Maybe just the last question, maybe kind of the reverse of innovation cycles and repowering. You made comments on that, which were appreciated. But it's a bit unclear to me, just given repowering is only happening now, but will, how TPI would be exposed to that theme, just given that my understanding was that the bulk of the repowering would be on older blades using legacy towers. Or A, is that true? And then, B, if it is, I guess, what molds will be used so that you can benefit from the repowering theme, given that these would be 40- to 55-meter blades?

Yes, I think the answer is actually a blend of those 2 points. There are some examples in repowering where blades in the 40- to 45-meter range, for example, might be used, and then others where the turbines are coming out with new turbines being installed. And so to us, Jeff, the brand-new turbines look just like new business even though what's important there is the gigawatts being installed are driven by repowering of something that was built 20 years ago, perhaps. So it's important even if load is not growing in a region as much, but if repowering is taking place and new machines are being put in, that helps drive demand for the new models, the new turbines, if you will. So there are some that are new turbines, and then some that are smaller blades. And we're doing some of both. Again, without getting into specifics by customer, we're doing some of both. I think we said this before, the majority of TPI's business is installing new machines, and that's still true, but there is a market support mechanism here for repowering that's helpful.

And, Jeff, as you probably know, some of it is site specific, right? There are some older sites that will be restricted for height and what have you. There are other older sites that are not. So in the sites where they're still restricted on height, it might be a blade -- a 40- or 50-meter blade. And in other locations where they're not restricted, you might see entire farms being pulled out and multiple turbines replaced by single turbines. So I think it's going to -- you're going to see a combination into that ladder case that would certainly support Steve's position that it really wouldn't -- we really wouldn't know the difference whether it was going to a repowering opportunity or to a brand-new wind farm.

As we mentioned earlier, this is not just a U.S. issue. Germany, France and Netherlands or other countries that are participating, some of those 4 countries are making up 83% of the repowering market over the next few years. So it's not just a U.S. driver, it's also occurring elsewhere.

Ladies and gentlemen, we have time for one more question, coming from the line of Joseph Osha with JMP Securities.

So we saw some recent news about the GE-Vestas IP lawsuit. And considering these are some pretty large customers of yours, do you see this having any impact on future backlog?

Yes, thanks for the question. So we've read the same news in the press that you all probably have. Again, we're not going to get out ahead of talking about any of this with respect to our customers' business. I don't think at this point that we see any specific risk to our business. But it's fairly fresh news, and I think we all need to just see how that plays out.

And then with China's continued expansion of the renewable sector and build out of high voltage DC transmission capacity to the city centers, could you comment on how these developments will be -- will affect continued growth within the segment?

Yes, I think in general, what's been really interesting from our perspective to see is some newer markets open up a bit. And the investment by utilities in insignificant assets would be wind farms and transmission lines, assets that they can then build into their rate cases and help to grow their business, help the utility grow the utilities business, if you will, through that work. So it's good that some of these recent projects, we're not really on -- in the market forecast or on the radar screen of some of the forecasters even a year or so ago. So it's helping to expand the market, and it's helping just fundamental pull from our customers' customers on the utility side.

Thank you. Ladies and gentlemen, that's all the time we have for questions today. I'd like to turn the floor back to management for closing comments.

So thanks, everyone, again, for your interest in TPIC. We look forward to updating you on our progress toward our stated strategy of global growth, customer diversification and expanded profitability, and look forward to seeing many of you at our Investor and Analyst Day in New York on Friday, November 17. Thank you.

This does conclude our teleconference. You may disconnect your lines at this time. Thank you for your participation.