Anders Runevad - CEO Marika Fredriksson - CFO

David Vos - Barclays Akash Gupta - JP Morgan Kristian Johansen - Danske Bank Casper Blom - ABG Pinaki Das - Bank of America Merrill Lynch Fasial Ahmad - SEB Marcus Bellander - Carnegie Alok Katre - Societe Generale Sean McLoughlin - HSBC Mark Freshney - Credit Suisse Claus Almer - Nordea Markets Jeffrey Vonk - Morningstar

So good morning, everyone, and thank you for calling in. Welcome to this Second Quarter 2017 Update. The usual disclaimers slide and then let me jump straight into the key highlights. Overall a solid performance in the quarter. Order intake increased almost 50% year-over-year and reached 2.7 gigawatt, and on the first half of the year we saw an increase of 12%. Revenue was 2.2 billion, a decrease year-over-year, but for the first half of the year basically on par with '16. Solid earnings with an EBIT margin of 12.6%. Also good to see a good strong performance in the service business both from a revenue point of view and from a EBIT margin point of view. And we have then launched a share buyback program of 600 million. As usual then, I will talk about the orders and the market, Marika will talk about the financials and then we'll come back to outlook and Q&A. So the overall regulatory environment I will say basically unchanged compared to last quarter and the trends to auctions and competitive tenders continue. We now, as expected, see it in all our regions and anticipate that that will be the new normal going forward. So the market is transitioning to an auction and competitive tender market. Looking a bit more in detail, in Americas no major change. We see the PTC structure is what drives the market in the current PTC cycle and of course the increase competitive of wind, as we have talked about before. And of course again I'm really satisfied with the order intake and the strong position we have in the U.S. market. And Latin America was probably the first region that started with auctions and we've seen good successful Vestas in Latin America. We expect auctions to continue; we will see in Mexico, Argentina, Bolivia, probably also Chile. And there is an announced auction in Brazil for Q4. Those have been postponed before, but that's at least what has been announced. In EMEA and starting with Europe, we've seen the first 2 auctions being held in Germany. The seasonal wind park has taken up the absolute majority of those auctions. In Spain in total for renewable energy 8 gigawatt has been auctioned and wind has taken about 4 of those 8 gigawatts. Sweden has done extensionally Green Certificate system, which is positive, and we've had a good order intake from there. And we also seen then auctions in markets such as Turkey and Russia. Very much the same, but of course from a much lower base in the Middle East, where we expect an auction in Saudi in Q4. China keep the 5-year plan and is executing on that plan. There exists curtailment in the market, but we also see good progress of addressing that curtailment. India also going to auctions, that creates some short-term certainty, but the 2022 target remains in place. And I will say for the broader Asia-Pacific region we see targets in place in most markets. Vestas is in this transition continuing to building our leadership position and the transition is of course fundamentally positive. It's a sign of wind competitiveness against other technologies and will longer term create a market that is less reliant on policy decisions, and therefore, reduce volatility. We're confident in our strategy and our financial ambitions. We know the levers; it is about levelized cost of energy and it's about operational excellence. And we will continue to leverage on our global reach, our scale and our technology and service leadership. So going a bit more into the details, as I said, order intake very, very good in the quarter, 2.7 approximately gigawatt, 49% increase year-over-year. U.S., Sweden, Argentina, Germany and China were the main contributors to the order intake and accounted for more than 70%. We saw a drop in the average price per megawatt in the quarter to 0.81. This is due to a mix of factors. Firstly, we have a negative currency effect in the quarter. Secondly, a fairly high proportion of Chinese orders, where the scope is very different from the rest of the world. Thirdly, in the quarter we had no turnkey orders, that of course again is a scope issue. And lastly, the dynamics of the market remains very competitive and especially with the increased number of auctions and competitive tenders. Looking at order intake also on a regional basis, you see Americas up 68%, very much driven by U.S. Also good activity levels in Argentina, offsetting decline in markets such as Brazil and Canada. In Europe or EMEA I should say, we see good activity level across many markets in Europe, but primarily then Germany and Sweden. The difference year-over-year is primarily the Statkraft order. In Asia-Pacific, strong development in China. We saw that in Q1 and we see that also in Q2 with a good increase in orders. And also good activity levels in India, Australia and South Korea. In the quarter, we have also further strengthened our product offering and Vestas is in a very good position. We are the only company in the industry with a significant volume and track record both on a 2 and what used to be a 3 and now is 4 megawatt platform. So on the 2 megawatt platform, we have announced new rotor sizes that will increase the production with up to 7%. And on the 4 megawatt platform then, we have increased the power rating as well as the rotor sizes with the potential to increase production more than 20%. And again, a broad portfolio covering all different wind conditions and markets. Looking at delivery then both in the quarter and for the first half year, we see a decline in delivery, it is basically in Europe and Asia. So Americas strong, very much also driven by the U.S. And in Q1, we of course also have help from the PTC components that was delivered in the U.S. In EMEA, we see good improvement and strong development in the UK, but also solid activity levels in markets as Germany, France, Turkey, but not compensate for the drop we've seen from South Africa, which basically is standing still for the moment, and also low activity level in Sweden. From a low level we saw a decline in Asia-Pacific. Here actually India and China are stable on their levels, but the decline comes from that we've seen very low activity from other markets in Asia-Pacific for the beginning of this year. Order backlog increased by €200 million equally between wind turbines and services and now stand at €20.2 billion, and that is despite a negative FX impact of approximately €600 million. The focus for the joint venture that we have on offshore together our partner Mitsubishi Heavy Industry, the focus for them in the quarter was to complete the first V164 project and get that really commissioned -- which it is now -- and also launching a new power upgrade of 8 megawatt platform to 9.5 and continue to look at a very healthy order backlog. With that, I hand over to Marika.

Thank you, Anders. So if we have a look at the income statement, we continue to deliver a stable Q2, although very tough comparison compared to Q2 of last year, which you clearly see when you make the comparison. But I would like to highlight that we delivered a very strong EBIT margin of 12.6% despite an impairment of a test facility that obviously have an impact on the P&L of €28 million. So all in all, very strong performance. And you can also see on the EBIT, although a decline in absolute numbers, it is a strong performance. And we also in this quarter delivered a very healthy gross margin, getting close to 22% in the quarter. The SG&A cost continues to be of high focus for us. And you also see when it comes to the volume and all the impact that we have from cost-out on the products that we continue to get a good leverage from our position when it comes to revenue. So we are down now to 6.7% here in Q2 of this year. So clearly in percentage a decline compared to last year. So very high focus on the SG&A also going forward. Service business, which is obviously a good mix in the quarter with a strong performance. So you see an increase of 14% compared to last year, which also have an impact on the gross margin obviously, and here we deliver a very good margin of 19.4%.I would like to remind you that we continue to see very high, good, stable margins in the service business, but you will also going forward see certain lumpiness. But it's definitely a good mix when it comes to -- for the overall company. And Andres talked about the order backlog increase despite the headwind we have on currency. Balance sheet continues to be strong and obviously something that we have been striving at and continue to strive at having a very strong balance sheet. So our net cash position increased to €2.6 million. And you have a positive net working capital development of 209 million. And I will come back to the moments. The solvency ratio remains within the boundaries of 30% to 35%, so we're close to 31% solvency in this quarter. So again a very strong performance on the balance sheet and I would say a control over the position we have and continue to have. And if we have a look at the change in net working capital, you see more or less the same pattern over the last 12 months and the three months. So what we have said in terms of using the inventory for -- instead of investing in additional capacity. We have done that for the last two quarters and that's also why you see a negative cash flow in this two quarters. That doesn't mean that we have changed our methodology when it comes to produce for orders. The underlying methodology remains. But instead of, again, investing in additional capacity, we utilize the inventory. So we build to a certain extent for inventory. But all in all, tight control over the net working capital over the last 12 months as well as over the last three months. So positive control over the net working capital. Warranty provision and loss production factor is just a reflection of the focus on the high quality of our products. So you see the loss production factor remains below 2%, so a very stable performance, although also here you will see certain fluctuations among the quarters. And you see that in Q2, we are getting closer compared to other quarters in terms of consumption and so we are getting closer to the provision. But it's all in line with the expectation and all in line with what we see as relevant provisions also in this quarter. The cash flow statement, again no surprise. We continue to have good leverage on our operating activities. So we are in that respect very healthy when it comes to generating cash flow from operating activities. You see a negative swing in the net working capital, again as expected, and that generates a negative free cash flow of 151 million. The cash flow from financing activities is obviously impacted by the dividend payment made earlier this year. The total investments, again solid. You see an increase compared to last year, planned for, a good reflection over the technology and the activity in the company. So we continue to invest primarily in the capitalized R&D, but also in molds. So obviously the molds have to be changed when we change technology, just to be very clear. Capital structure, good performance also on the capital structure. You see here we are net debt-to-EBITDA well in the territory or well below continuously. So we are negative 1.4 billion. So obviously the focus remains. Solvency ratio, also here solid performance, well within the boundaries and slightly up compared to last year, but more or less in line. The share buyback program that Anders announced earlier is 600 million. And also bear in mind that we actually did a share buyback earlier this year of 100 million and we also paid dividend, so we are in the 1 billion range when it comes to return to the market. And there's no changes in our dividend policy and the priorities for capital allocation remains unchanged. The return on invested capital obviously more than double-digit at this point, so very healthy levels and driven by the strong performance on both income statement and the control over the balance sheet. By that, I give the word to Anders.

Thank you, Marika. So outlook, we maintain the outlook for 2017, which means revenue between 9.25 billion and 10.25 billion, EBIT margin between 12% and 14%, total investment approximately €350 million and a free cash flow of minimum €700 million. Also worth noting is that it's based on current exchange rates and previous outlook of course based at the exchange rate at that time -- and since then we have had headwind on the exchange rate, primarily U.S. dollar to euro. So with that, we go over to Q&A.

Thank you. [Operator Instructions] Our first question is from David Vos, Barclays.

Could you please on the U.S. firm up what your deliver schedule looks like for the rest of the year? We've asked that question at Q1. You said you needed a little bit more time for that. Has that time now passed and can you therefore be a little bit more specific about what you think in the U.S.? And also could you comment on -- with respect to order intake there how you are tracking versus your expectations going into 2017? And then the second question would be, if you could specify whether you can still offset the observed price erosion in the market with your usual mix of cost-out, productivity and engineering that you have been able to do in the past?

Let me start then with the U.S., I think overall of course really happy with the performance of the U.S. as we were also during the end of last year with the PTC components, but also what we see begin of this year when it comes to taking the orders on that potential. I think when it comes to delivery, I mean of course you never know what's exactly foreseen through the year and what comes out of the year. But I think from a visibility point of view, of course we do orders that we have now taken. We feel good about the U.S. market and our position in the market. But I will not get into any sort of precise delivery schedule. When it comes to the other part, there is of course the potential of PTC components for 2021 and there we have definitely started discussions with the potential customers, there is an interest in that, but it's too early to speculate on how that will pan out. And honestly, I don't think that we will know that for sure -- experience of previous PTC cycles -- until the very end of the year. The other question probably a bit on how we see PTC components on order compared to continuous construction, I think it's fair to say, as expected, the majority of the early orders are around continuous construction. But also fair to say that we see some PTC components order also coming in. On your other question on how we handle the market and the competitive situation that we see in the market, the levers we have in place as before, it is very much about focusing on levelizing cost of energy. And that of course are 2 main components in that. One is the technology and that is more generation from new technology as we have also then launched this quarter. So it is about the product. And the other big part of the question is of course the cost-out that we will continue to drive. So I think we are familiar with the levers to continue to increase our competitiveness in the market. And of course we made -- and how it will influence margins, I think what we say there is of course that we maintain our margin outlook for this year.

Our next question is from Akash Gupta from JP Morgan.

My first question is on the price per megawatt development. I mean, if you look at the headline numbers, then obviously it implies down 9% year-on-year. But you have cited various factors behind it. What I am after is that if we strip out these factors and look for like-for-like price development, then maybe if you can help with that, because your U.S. competitor reported that it was like flat pricing in their renewal orders. So that's question number one.

Yes. I mean, it's actually very hard to single out. It is a mix of factors, as I said. I mean, what is maybe a little bit easier to single out is of course the currency effect and that is 0.02 I think in the quarter. But then it is really the factors that I talked about. It is about the scope and in this case then mainly two parts of the scope, which is a higher proportion of orders from China, which -- I mean, rule of thumb before has been that megawatt there from a scope point of view it's about the half the value of a megawatt outside China. So the scope is very different there. And the other part is that in this specific quarter then was that we had no turnkey orders. And as I also said lastly, we see an increase competitiveness in the market, very much driven by the auctions and the competitive tenders. That's also nothing new, as I said. I mean, we've of course seen auctions in parts of the markets before, I would argue very competitive tenders in the U.S. and an auction like system especially in the PTC cycle we are now that is a longer cycle. But what's different now is of course that we see this now in all different markets. So exactly on the 0.8 year-over-year what are -- what's in this different buckets very hard to say, but of course they all play a role.

And my second question is on the guidance, which is still €9.25 billion to €10.25 billion wide. So I was expecting maybe you would be narrowing the guidance at this point of time, but you have left it unchanged. So two questions here. One is on, do you expect any -- maybe some revenues moving from this year to next year? Is that the reason why the guidance is so wide? And secondly, does this guidance factor in announced orders so far in third quarter at all?

So the guidance we have -- well, we have done the similar analysis as we do and the similar forecasting and simulations as previously. And just to highlight that nothing new actually for us, is that we have a busy second quarter, but that also means that we have a lot to deliver. And a lot of it is -- you also see a currency headwind at this time; that also taken into consideration when we talk about the low range of the guidance. And we also have a lot of deliveries late in the year that actually have to be performed as expected. And there we would have the regular weather, crane and everything else that could potentially go wrong. So no differences in how we have made the analysis. We're still striving for the higher end as we always do.

Our next question is from Kristian Johansen from Danske Bank. Please go ahead.

So my first question is on the infra-out orders. So the order intake has been quite strong. Can you just give a bit of flavor on the level of infra-out orders you've received this year versus what you expected at the beginning of the year?

Okay, on the order side in terms of the infra-out, you would see normally in the earlier part of the year is when you see most of the infra-out orders. But having said that, you would also see that some of the U.S. PTC components would qualify for infra-out to certain extent. And the portion of infra-out in this particular quarter is high or fairly evenly spread. China could be one of those countries, for example. So nothing abnormally high or abnormally low at this point in Q2.

So in line with sort of what you expected at the beginning of the year?

Yes, that's what I'm saying, Kristian. Yes.

Okay, great. Then my second question, obviously there's a lot of focus on the impact from auctions here. As you mentioned, Latin America has sort of been a first mover with auctions that's been around for quite a while. I mean, when you look at sort of the average budget margin for your Latin American business where auctions has been around for a while, is there a notable difference to the rest of your business?

No. And that's a little bit what we have said also. Even if -- I mean, you would have a low China scope. You would have a fairly low U.S. scope. You would have a higher India scope, for example. All in all what is important for us and as Anders have highlighted when it comes to scope is obviously the price per megawatt. But having said all of that, the request even if we have very little in Q2 when it comes to EPC contracts or nothing, we see that all in all that is increasing a little bit or the demand for EPC contracts is increasing. But the margin is not related. Obviously, when you have a higher scope, you have more parameters to work with. But all in all, the margin is fairly similar, independent on the scoping of the projects, because you also have a lower cost on lower scope.

So despite the average price dropping in Q2 on order intake, we shouldn't expect a similar impact on the margins?

Correct.

Our next question is from Casper Blom from ABG. Please go ahead.

My first question is regarding your payout of cash towards -- you're paying out roughly €1 billion this year, which I think on a normalized basis is roughly also your free cash flow given the timing differences around New Year this year. Is this also sort of a fair assumption going forward that you now have the balance sheet you want and you can more or less payout your entire free cash flow? That's my first question please.

Well, what we have said is that the dividend policy remains and -- I mean, that's a 25% to 30%. The share buyback will be -- the cash that we have or if we have any excess cash during the course of the year, we will obviously invest in the business first hand and we will also do bolt-on acquisitions, as we have said. Haven't found anything at this point obviously. And then we have said also that then we will revisit whether we return to the market, and in this case, we thought that the timing and the sizing was appropriate. But we are not committing to any certain level because obviously we want to have and continue to have the strong balance sheet we have at this point.

Fair enough. But it says that you're comfortable with the balance sheet you have then today I suppose. You can decide whether you want to comment on that. My second question is also a little bit about prices here in the quarter getting a lot of attention. Do you have sort of -- when you participate in these auctions, where, as Anders said, things are getting very competitive, is it your impression that you are sort of getting a higher proportion of orders today than you did before the introduction of auction systems on the back of competitive levelized cost of energy? If you can talk a little bit about the -- maybe the dynamic of these available CE really coming in the spotlight in the introduction of auction systems.

I think -- I mean, it's very dangerous to generalize, but I think of course there are some common themes on the auctions. I mean, if I look -- as I said, I will say that we've seen auctions for quite some time. The difference now is that we actually see it in all regions. And if I look at our performance the last year, the last 2 years with auctions and competitive tendering, as I said, in the U.S., very similar to an auction kind of system, where you have a longer period usually to deliver on because they are a bit more forward, not -- I mean, maybe Brazil is the most extreme example where there has been some auctions that are called A minus 5, which means that the customer bid for an end of the PPA 5 years from now. So they are a bit more forward looking in that even if Brazil maybe is the most aggressive in that. So that I will say is very characteristic of the current PTC scheme we are in as well. So having said all of that, I must say I'm really confident with our position when it comes to growing faster than the market during these last 2 years that I have talked about. So I'm confident in the competitiveness of Vestas. But then having said that, of course there are also differences between auctions. I mean, sometimes it's differences in timing. Sometimes there are high requirement of local content. Sometimes there are other type of local requirement, which you of course also have to take into account. And that means that the levelized cost of energy from the turbine as such is the percentage that that plays in to the overall PPA price that is then bid in the auction, can vary quite a lot dependent on the different other requirements that are in place in order to participate in the auction. So a generic answer, but also with the caveat it's very hard to say that all auctions have the same criteria.

If I can just follow-up on that, you mentioned that that some of these auctions have very long lead times. Is that also sort of a part of the explanation why there sort of seems to be a general impression that the prices come down in connection -- together with the auctions basically since we are talking about technology that will be available 3, 4 years from now, where levelized cost of energy is expected to be lower and hence people can also bid a lower price?

No, that's correct. I mean, if I mean, very, very simply put it in a fade in and probably a little bit simplistic, but I mean to take the extremes. I mean, in a very stable feed-in tariff environment of course we and the rest of the market we sold the product we had on the shelf, and if that qualified with the PPA within the feed-in tariff, the project was being built. And if you then take the other extreme, as I talked about, probably Brazil, then that has an auction that is -- where the customer then bids for a PPA that is 5 years from now when it's going to be built. Of course if we then use the current technology, you will never win that auction obviously. So of course there you have to partner early with the customer and anticipate the levelized cost of energy to some extent on the technology you have. I mean, those are the 2 extremes. And then you have everything in between that when it comes to timing, and as I said, different other qualification criteria into the auctions.

Our next question is from [indiscernible], Handelsbanken Capital Markets.

If you just can follow up on the last comment you made here, because in an event like you mentioned, for instance, in Brazil, if you commit 5 years, 3 years down the road together with a partner, what kind of commitment do Vestas take here and what are the risks, so to say? First question.

I mean, for competitive reasons of course I will not go into exactly how we partner with our customers and what kind of commitment we are doing. It's, as I said, a very competitive market.

So you need to make a commitment of some kind.

No -- of course. I think we of course all have to bid the technology that we feel confident we will have at that point in time.

And then a more upstate sort of question relating to the strong sort of say pipeline that you have of 4.3 gigawatts which you mentioned is being under completion. How should we think about that in the coming quarters? Is that more or less also be delivered throughout second half and how will it be split between Q3 and Q4?

If you look at the megawatt or the projects under completion -- are you asking in terms of how we see the -- you're looking at the completion for -- on our inventory or…

Exactly.

Yes. So there you have to note that also we have the turbines booked under inventory that is supply only. But then you also have turnkey project on top of that that would sort of have an impact on the P&L. So it's hard to sort of make a one-to-one. But the underlying sort of methodology when it comes to what we -- how we book and what we book remains unchanged. But you cannot make sort of the match in that perspective. But we obviously are busy, if that's sort of the underlying question.

Or I guess the majority of the 4.3 gigawatt will be delivered throughout second half.

Yes. And that's a fair assumption.

Our next question is Pinaki Das from Bank of America Merrill Lynch. Please go ahead.

I'd like to follow up on this megawatt equivalent on the completion. Even that number is up significantly compared to last year, you're at 4.3 gigawatts now, at this stage last year you were 3 gigawatts. I just wanted to understand when you look at the second half, what has changed that you have a much higher inventory number although you're not expecting higher volumes year-on-year in 2017 versus 2016? Why is the inventory so much higher now than at the same stage last year?

Fair question, Pinaki. And basically what I said is that in terms of activity level we can decide obviously to invest in overall capacity and we can decide that we use our position when it comes to networking capital and thereby increase the inventory. And that is basically what you have seen this year and that is a deliberate choice. And the activity level, if everything goes according to plan, will be high definitely. But you also have to bear in mind even if the vast majority is for activities '17, you would also have some activity entering into '18. So it's sort of it's not everything in '17. But then obviously the higher activity level also means certain uncertainty, and that's why we have made the simulation and have the similar view as we had beginning of this year.

I mean, you're almost suggesting that there's some sort of capacity constrain and therefore they are investing in their inventories now. But as we know that '17 is a slightly lower year than '16, there should not be that much of a capacity constrain, no?

I mean, overall what we have -- if you see how much we have increased our capacity without making any major investments, we obviously optimized the usage of our capacity and that's also why we -- and at this point we can afford to build inventory to meet the overall demand. But what I'm saying is that not everything is for '17. We can also -- some can ship into '18.

And my second question is around your FX sort of headwinds and you mentioned that obviously in your guidance. You hadn't -- you obviously didn't factor in whatever FX might change over the year, which is fair enough. And you also have the backlog, which is now been adjusted down by FX. What I wanted to understand is, most of the PTC components have already been delivered in Q1, so that FX is kind of already gone. So you would have still some order which you booked maybe end of last year or in this year which will probably be delivered in Q2 and also Q3, Q4. And what is your strategy around -- you hedged the FX on the orders that you've already booked. And going forward when you look into new orders that you might book, is FX an issue or is it more like a -- it's an ongoing thing where you will just -- the pricing will be slightly different because of FX?

I see where you're coming from Pinaki. Again, all in all, we try to be as naturally hedged as we possibly can and that's where obviously the global industrial footprint and also the sourcing capability that we have is extremely important. But in -- when we talk about -- the currency effects and the headwind on the guidance is primarily translation impacts, so you don't see the transaction impact. But at any time if we don't have sort of the perfect hedging position, then we hedge the specific project, but that we don't do until we have a firm order intake.

But again sort of reconfirming on that, when you had booked orders in the U.S. end of last year for delivery in 2017?

That has already been dealt with in terms of..

Yes. You've already hedged the FX on it?

Yes. When we booked the firm order intake, yes, then we hedged the contracts.

So you basically have a negative impact on your backlog, but you might have a separate positive financial item somewhere. Is that the way I should think about it?

Yes. I mean, the translation impact, yes, if you have a negative impact from the U.S. dollar, you would see certain positive impact on in particularly fixed capacity cost. But that would be fairly minor as we still have a big euro base.

But do you have a financial asset in front of it? Like do you have an FX contract which in the positive territory now and you have a negative impact in the backlog?

Okay, do you mean the contract -- I would say that's fairly insignificant. It is -- all in all, when it comes to the backlog, it's the translation. And also when you look at the forecast going forward.

Our next question is from Fasial Ahmad from SEB. Please go ahead.

Two questions from my side; firstly, production in second half. Will you continue to produce at full speed in the second half of the year? That's my first question please.

Yes. I mean, we will continue to produce on full speed or sort of current level. And then of course -- I mean, as you know, we have production facilities in all the major geographies and of course -- and that's a little bit coming to the last question as well, that of course we have an overall production capacity and then you have individually regional capacity production as well. And of course the advantage of being global is to try to always match the utilization with the demand in the market. But there is of course a physical limitation a little bit depending on how the market goes and how much you can optimize that. But I mean there is no slow down in manufacturing. And if you look at blue color workers, they are fairly constant.

And how should we be thinking about the inventory level by the end of the year? Normally you draw down on the inventories in the second half of the year. How should we think about that in 2017?

I mean, obviously we will utilize the inventory at the end of the year. But then exactly what position we would have, we're not sort of commenting on, but that will be a normal pattern, yes.

And then just a question relating to guidance, a lot of questions relating to your top line guidance. But when I look at your margin guidance, the lower end of the 12%, can you explain why we should not see the normal kind of seasonality for margins in the second half? I mean, you're essentially indicating flat margins in the second half compared to the first half year. Why shouldn't we see a positive impact from leverage? Or there's some one-off costs or is it pricing which is hitting your margins in the second half? If you could elaborate on that please.

Well, all in all, if everything goes according to plan, yes, there would be a very high activity level. But that also means that you -- that your suppliers works perfectly, that all the logistics works perfectly, that you get all the cranes that you have ordered and that you don't have any snow storms at the wrong timing. So this is -- when we make the analysis and do the simulations, this is the considerations we have. And that's why we have chosen to keep the guidance, because there will be a normal uncertainty. And then on top of it, you have currency headwind at this point when it comes to translation.

But no one-off which we should be factoring in in the second half of the year?

Not that we foresee.

Sorry, ma'am, I didn't understand that.

No, we don't foresee any major one-timer effect in the second half.

And then just one final question. You're launching a number of new products for your 4 megawatt platform. Should we be expecting any significant ramp up in R&D or tangible CapEx from that?

I mean, we feel very comfortable with the efficiency we have on the CapEx level, i.e., capitalized R&D in that respect. And I think we have proven to be very efficient both on accommodating new technology and capacity. I mean, the guidance that you see this year, we feel comfortable with the level that we are at. And then obviously you would have an impact from activity level primarily when it comes to molds.

So the €450 million level annually, we can also extrapolate that to '18, '19 due to these product launches?

Well, with the -- as I said, the CapEx level that we have is something that we feel comfortable with.

Our next question is from Marcus Bellander from Carnegie.

My first question is also regarding the guidance and the FX effects. If you could quantify how big of a negative impact the currency movements we've seen thus far have had on the top line or will have had on the top line by year-end?

Well, most of the negative headwind will come in the forecasting part when it comes to currency and that's again translation. And what we see is around 2% to 3% impact from currency.

And then the second question, the 28 million impairment charge, could you just provide some more detail on that?

Yes, absolutely. That is the 28 million impairment is a reflection of usage of a test facility and therefore -- as the usage of this particular test facility we assume going down and that's why we have impaired by 28 million. So that is a reflection of how much we anticipate that we will use it going forward. And then obviously we have other test facilities that we're moving the testing to. It depends on the type of turbine obviously.

And stupid question maybe, but does this have anything to do with the burning turbine in Denmark?

No, absolutely not.

Our next question is from Alok Katre from Societe Generale.

Maybe first one just following up on the CapEx and the whole inventory production capacity sort of issue. Are we sort of getting into a stage where capacity is starting to become a constraint and is that sort of partially reflective of the expectations of how the U.S. sort of market will pan out, the high volumes in the next few years and then obviously we don't really know how it pans out later on? And then of course related to that being, some of the new markets, Russia, Turkey et cetera, seem to be insisting on a high level of local content that requires factories. So should we sort of expect CapEx to creep up in terms of percentage of sales going forward? So that was number one. The second one was, if you could just comment on developments on the end markets and particularly U.S. and Germany? U.S. of course some of your peers slightly more of a push out of installations from perhaps 2018 to '19, '20 due to tax equities issues. And then Germany, where of course the whole community wind farm -- unpermitted, let's say, community wind farms wining the auctions seem to be creating some sort of a vacuum potentially in latter part of 2018, early 2019. So any comments over there would be useful.

So if we start with your CapEx questions, I think that -- as I said earlier, the overall platform that we have is extremely useful also not only when it comes to production, but also sourcing. And the increases we have seen when it comes to capacity within the range that we have guided for this year, we are very comfortable with. We have also said that as we have a very strong working capital position and a very efficient one, we will when we see fit use the inventory or use the position to actually being able to build inventory without having a major impact on the overall position that we have on net working capital and cash flow. So there's no changes. What I would say that when it comes to local content requirement, the global footprint and the know-how that we have to enter into new markets both when it comes to production, sourcing is extremely valuable. And also on that note, we are extremely efficient when it comes to accommodating any local content requirements.

Should that drive higher CapEx, I mean if you have to build factories in Russia or Turkey?

And that's what I said. As we are extremely efficient in the use of our CapEx and also moving molds if necessary because they are movable, we're talking about lower level of investments and so we don't see any changes in our CapEx requirements going forward.

Okay, let me try to see if I can answer your other questions. I think on the U.S. tax equity, there are no news in this quarter compared to what we said in the last quarter. I think it's of course, to start with, extremely hard to speculate about the U.S. political situation and the possible tax reform to start with. And of course there is a theoretical then possible impact on the tax equity, but I think that is of course a theoretical impact. So if I look at the market again, if I look at the order intake we take in the U.S. and if I look at discussions we have with our customers, we have not really changed our view or the market size of the U.S. in the current PTC cycle. Germany then, I think that is a good question. I think it remains a little bit to be seen. The difference between the seasonal wind park and the non-seasonal wind park is of course, as you say, that the seasonal wind park have a longer time before they have to complete the projects and they don't need a building permit to participate in the auction. So there is potentially a longer lead time on seasonal wind park projects compared to non-seasonal wind park projects. So how that will pan out -- and of course the two first auctions now the absolute majority went to seasonal wind parks. So I think how that will pan out will of course -- depends a lot on their internal time schedule. A negative there is that they don't have building permits. That will point to a longer process. The positive is of course -- like always time to generation is always positive for your business case if you are -- if you have a project. So of course there is a natural tendency to try to get the projects done as quick as possible from a time to money point of view. So I think it's very hard to see how these two pans out. But from an overall timing perspective, no doubt that they have a longer time to completion. And then I think there will be one more auction with the current ruling and then the auctions after that will -- there will be no preferential treatment for the seasonal wind park.

Are you seeing a push out of the U.S.? I mean, I can appreciate that the overall market size view over 2017-2020 hasn't changed, but just the phasing within that is it -- would you agree with some of your appears saying that the phasing within the 2017-2020 timeframe is shifting out a bit more towards '19-'20 rather than '17-'18?

I mean, what we said before was -- and I think now the overall market that we expect '17 from a delivery point of view to be lower than '16. I think -- and that we maintain. I think then it's very hard to predict. Again, if I look at it more short-term and order intake and the projects, we're all firming up. No, firm orders for the first half. I feel fairly good about the position.

Our next question is from [indiscernible] from Macquarie. Please go ahead.

I just have a follow-up on the last question around the German market. So, Anders, do you think next year the German demand for 2018 will be quite bleak given the sort of push out on the citizen based projects as well as the fact that the feed in tariff projects awarded in 2016 are most likely to be commissioned in 2017?

I think, yes, likely from a delivery point of view of course '18 will be down on '17, but exactly how much, I think it's very hard to predict now. I mean, that we will get clarity on once we start to see orders firming up around this seasonal wind park. But of course there will be a point in time where the market goes to the sort of -- the auction volumes that has been decided, and exactly when that will happen, I think it's hard to predict, because of what I said then, they have a longer lead, they have a possibility of a longer lead time. On the other hand, of course they have an incentive to get to generation as early as possible. So, yes I mean, that's of course something we have to stay close to and understand better now as we start to see those auctions turning into orders.

And my second question is about Mexico. So that's obviously not been a huge market for you guys. So what was the main drivers of winning the recent 424 megawatt project? Did it come down to price? Was it technology? Was it the 4 megawatt platform?

Which market did you say?

Mexico.

The Mexican market.

I must say that we actually have a fairly good position in Mexican market already from before. So I will say that our market share in Mexico is probably on line or maybe slightly below our global average. I mean, Mexico always been a solid market for us and of course we are happy for the order.

Our next question is from Sean McLoughlin from HSBC. Please go ahead.

My first question, just coming back, Marika, to a comment you made earlier about demand for EPC contracts increasing. This probably will seem to square with the argument that lower scope impacted your order pricing in Q2. So I was just wondering can you specify which regions, which markets are looking at more turnkey and how you expect this trend to impact your pricing going forward.

I mean, normally we would have a maximum around 5% of EPC contracts. But we clearly see depending on markets but I will say a very typical turnkey and EPC market would be India, for example. You also see other markets where you see certain requests. We didn't have anything in this quarter, which obviously has an impact on the price per megawatt, as Anders highlighted. I mean, this is not blowing anything out of proportion, but I think also from a competitive perspective that you can do any type of project or project is extremely valuable when you have a global footprint as we have. I mean, obviously the customers comes from anywhere in the world and we have the possibility to provide any type of contracts and scoping of contracts.

I mean, thinking as well about the buyback and the lack of M&A, you mentioned yourselves bolt-on acquisitions, I'm just wondering if in the context of that you could give us any further thoughts on your strategic aims to broaden your product offering into storage?

As I said, I think our strategy remains that -- and I mean we have done some bolt-on acquisitions, as you know, in the service business. I mean, if we find the type of bolt-on smaller acquisitions, we will of course look at it. But also clearly stated as before, we don't see any attractive big type of acquisitions, consolidations for us. We are -- but we are looking at -- as we also talked about before that we see an interest in the market for hybrid type of systems, it's from a megawatt point of view currently fairly small projects, but we think very interesting projects for the longer run. They are quite often a combination between wind, in some cases solar, but almost in all cases some kind of storage or increased capacity factor by new technologies towards the grid. And that could be a different kind of story. It can be pumped hydro. It could be batteries. It can be pure enhancement on how we handle the grid connection. And those projects we think are very interesting. Again, our value-add in those project is of course to take the project responsibility, it's to invest and look at the technology of energy management, it's what kind of gains we can do on the wind side with an enlarged project. But it's not sort of -- we are not looking to acquire a battery company and start to manufacturing batteries or a PV panel company to start to manufacturing panels. We think that our other company is more suited and has the scale on those components into those projects.

Does that also mean that there's also an internal R&D ramp in this area?

No, we have actually had -- as I think I have talked about before, we have an innovation part of our organization that are looking at a lot of different technologies. I mean, when we see a bigger interest in a certain area, we move the resources into that area and try to be prudent, as usual, of keeping the overall frame.

Our next question is from Mark Freshney from Credit Suisse.

I have 3 questions please. Firstly, one of the things you didn't mention for the lower pricing in the backlog is the power mode. I'm just wondering if that's important, because most turbines now come with this override stroke power mode. Secondly, on the FX rates, if we put aside all of the financial hedging that you do when you win an order, is it fair to say that you have had a large competitive advantage over the last 3 years in the U.S. as you've being able to ship sometimes complete systems from Europe into the U.S. for some of the 3 megawatt projects and basically manufacture in Europe 20% cheaper than you would in the U.S. So if you strip away hedging, there seems to be a big competitive advantage that you've had. And thirdly, just on the auctions, we are now in a system where auction prices are basically the market price for power. I'd be interested to hear on those, whether you think the days of competitive auctioning are actually coming to an end and we'll move to a system where auctions won't matter or won't take place and how that would impact your business.

I mean, your first question, the power mode definitely also is a factor in what we discussed. And as you sort of rightly point out, of course if we, for example, had most 3.0 megawatt in our back -- in our order price per megawatt a couple of quarters ago and now have mostly 3.6., of course that has an impact. And of course we can always discuss and that's why it's so hard to say exactly what is what. And is that then a pricing issue or is it a product rating issue? That's of course what makes the discussion a bit more complicated. I mean, on a like-on-like basis, of course that we have higher power rating and therefore more megawatt, that of course has an impact. I mean, I will say -- on the U.S. question, I will say a bit the opposite. I think the big competitive advantage that we had during the last 2 years in the U.S. is actually our U.S. manufacturing. And if you look at the absolute majority of the volume we have sold in the U.S. and a very big part of what we expect to do -- let's see -- is on the 2 megawatt platform that we have fully localized in the U.S. And I think that is actually an advantage. And we are currently in the process of also -- because the good thing is that on the manufacturing side in the U.S. and on the blade side, we can also do the 3 megawatts. We're actually looking at localizing the 3 megawatt now as we have started up that work for the U.S. So that I think has been a key for us, 1 of the keys for us in the U.S. market. On the auctions and market price, I think that -- I mean, as I said, we are in a transition. We are reaching -- or wind is reaching what you can call grid parity, even if I don't really like the word, because it's very hard to define, because it -- there will always be policy in the energy industry and that goes for renewable and that actually goes for fossil fuel as well. So I think that -- but of course -- I mean, once we reach the point where we clearly see no insight, it's possible that there will be another type of system. But for me, that will have many similarities to what we now call the auction system. It will be then based on the electricity need in the different counties. And there dependent on how the setup of the energy market is, you can view it as an auction or just a tendering for capacity. I think the key question there for me is that if you do an auction, you have to have something that you auction out. And I guess that is a little bit what you are thinking as well. So that is then that you either have a monopoly of power in the market or just 1 grid operator. But you have to have of course in an auction something that is auctioned out. Otherwise it's, as you say -- I mean, it's a market where it's just sort of built when need arise. But since the energy market overall in the majority markets are policy, there is a policy, I think you will see this type of system where you plan for additional capacity and there is some party that sort of do the overall capacity planning and therefore it will be an auction like system.

Our next question is from Claus Almer from Nordea Markets. Please go ahead.

Just one question from my side. There has been a lot of questions regarding all these auctions and the impact on pricing of course. Can you put a little bit of flavor on the quality of the backlog, should we see the current backlog being of equal quality as what you're going to deliver in 2017? That will be the question.

Well, I would say what we have said earlier. First of all, we have a very high quality on the order backlog and we have no intention to lower the quality of the order backlog. And then how it will pan out exactly -- but also bear in mind that you have anything from 20 -- or 6 months to up to 24 months. And when and how that potentially will increase in terms of deliveries because there is more forward sell, then we have to come back to that. But all-in-all, it's a very high quality on the order backlog.

So that will be the same answer when we asked this question in the past, where you normally, Marika, said, well -- yeah, the quality is equal to what we have seen being delivered.

But that's a little bit the point also, Claus, when Anders talked about the price per megawatt, because you would -- you don't necessarily see a correlation between the price per megawatt and the margin of the projects. And obviously when I talk about the quality of the project, it is a margin question as well. So we take good healthy projects. That is the -- that's the methodology.

So when I talk about quality, that will be an on project margins of course.

Yes. And that's what I say. I mean, there's no correlation on price per megawatt and the project margin.

Our next question is from [Lars Kehl] from Nykredit Markets.

Also a question related to these auctions. And, Anders, did you say that you didn't expect all these auctions to have a negative impact on your margins going forward, did you say that before? That would be my first question. And secondly also related to these currency movements, would you indicate just roughly how big of a percentage of your sales is in U.S. dollar or U.S. dollar related currencies? That would be my questions.

No, but we have not started to guide for margins further on than what we normally do. So we keep our guidance on the margin for this year. Overall margins we have to -- we will follow our normal process for -- when we do the next guidance. So what we talked a lot about of course was that there is a competitive pressure in the market. There are a lot of levers on competitiveness and price is one of them. But as we have talked about, in our eyes there's a lot of other levers as well. When it comes to the margins, we feel that we also know the levers and also the levers we are working on. But I have not given any indefinitely promise on margin developments for Vestas, so I mean we will follow our normal process there on how we guide for our EBIT margin going forward. What was the second question? U.S. dollar.

U.S. dollar, okay, the order of magnitude of the U.S. dollar. All in all I would like to highlight again that our assumption is to be as naturally hedged as possible. And when it comes to U.S. dollar, all-in-all -- I mean, obviously the U.S. market is a U.S. dollar-based market and that's a good portion of our revenue and there we are more or less 100% naturally hedged as we have a big operation in the U.S. And also what I would like to highlight again is that it is primarily as we report in euro, the primary currency impact that we see is translation. So it's not transaction as we try to be very naturally hedged. Then exactly how much is U.S. dollar-based, I would go for the U.S. market, that's where we primarily see it -- and some on the emerging market, but the vast majority of the U.S. dollar base is U.S.

There's room for one more last question.

Our last question is from Jeffrey Vonk from Morningstar. Please go ahead.

I have two questions if I may. The first one is going back a little bit to the deliveries. We have seen lower deliveries in the quarter, but very strong and high manufacturing activity and I think a lot of things has to do also with the transfer of risk. Can you give a little bit more color on that? Should we read something into that? That's my first question. And the second question is also coming back a little bit on the share buyback. I appreciate a point which was raised on the earlier question regarding the flexibility. But can you please come back a little bit on, let's say, your decision to go for share buybacks instead of raising the dividend, because in my opinion dividend can be raised given the fact that your profitability has increased on a structural basis the last couple of quarters?

If I start with the last question. And I think, we are happy with the policy that we have outlined, our way of looking at share buyback, the share buyback we have decided this year and of course the policy that we have also outlined of when we think it's prudent to maintain the balance of having a strong balance sheet, which we continue, I think is very important, and when we see room then for a share buyback. Dividend policy is of course at the end of the day a question for the board. But as I said, from management we are definitely happy with the policy we have. When it comes to the activity level also for the second half, it's as I think Marika explained, we are a project business. We see timings as you say, when we recognize the revenue, so the completion of the projects, which of course we play a big part of, but also our customer. And also we are then dependent on normally very high activity levels towards the end of the year. And to do the revenue recognition, we have to look at things also outside our control. We have mentioned all the normal weather, seasonality, supply from subcontractors. And then we also have grid connections from the customer and things like that that the customer has to be ready for in order to do the final revenue recognition. And as usual, we have uncertainty of that timing towards the end of the year. Okay. With that, again a reminder of Q3, that we will have that on the 9th of November. And then I just want to thank you for calling in, thank you for your questions and thank you for your interest. And I'm sure I will see at least some of you during the next couple of days. Thank you.