OSRAM Licht AG

Andreas Spitzauer – Head-Investor Relations Stefan Kampmann – Chief Technology Officer Olaf Berlien – Chief Executive Officer Ingo Bank – Chief Financial Officer

Peter Reilly – Jefferies Sven Weier – UBS Lucie Carrier – Morgan Stanley John Quealy – Canaccord Genuity James Moore – Redburn Uwe Schupp – Deutsche Bank. Alok Katre – Societe Generale Guenther Hollfelder – Baader Bank

Good afternoon, as well as good morning ladies and gentlemen. My name is Andreas Spitzauer, Head of Investor Relations of OSRAM and I want to welcome you to OSRAM's conference call for our third quarter financial results 2017.

As a reminder, the conference call will be recorded and is available on our homepage www.osram-group/investorrelations .com. You can find today's presentation there as well.

It is now my pleasure to turn over the call to the Dr. Olaf Berlien, the CEO; Ingo Bank, the CFO; and Dr.

Stefan Kampmann, the CTO of OSRAM.

Yes, thank you very much, Andreas. Ladies and gentlemen, also from my side a warm welcome.

Let me start with the financial highlights of our third quarter. Thereafter my colleague, Ingo, will provide the more detailed explanation of the financial figures and of course it will be followed by a Q&A session from all of us.

Looking at Slide 3, you can see that from the business point of view the last quarter was successful for OSRAM. The financial results show that we have maintained our strong performance and we managed to push further ahead with our innovation topics.

Revenue rose on a comparable basis by over 3%. It is important to note that revenue in the third quarter of last year were supported by pull-forward effects ahead of the carve-out of it.

Excluding these effects our operational revenue growth in the third quarter of this year would have been more than 8%. At the same time our adjusted earnings margin remained on a high level of more than 16%.

This solid development was driven by ongoing high demand for our LEDs and by the automotive business. As we move to Slide 4, overall, we are performing very well in the automotive business and we are gaining market share, especially in China and the rest of Asia.

This is reflected by the figures on Slide 4. Despite the fact that the market development in North America is more challenging, we have been able to grow due to our strong market position.

At the same time the situation of the construction industry in North America remain demanding. So in our Lighting Solution and System business the challenges persist, however measures to improve profitability are being implemented.

Just over 2 weeks ago we reached a reconsolidation of interest and signed a social plan for the main side of our Luminous business in Germany. The goal is to reduce the workforce there by 1/3 until 2020.

And as you can see OSRAM is performing well. One of the main drivers is our technology leadership especially in the automotive sector; here we are a key player.

As you know, we are expanding our worldwide LED production capacity in order to meet the high demand. We are fully on track and partly ahead of schedule.

In March for example, work began on the expansion of the Regensburg site. To meet the high demand we even invested additional millions of euros to be some weeks ahead of the schedule.

In 2 weeks, we will start expanding our back end capacity in Wuxi in China, end of November we will open our new chip production plant in Kulim in Malaysia. So ladies and gentlemen, let's move to, switch, Slide 5.

OSRAM is also laying the basis of the future success of its business. Only recently we acquired a 25% stake in LeddarTech.

LeddarTech is a Canadian software company that develops innovative solutions for the automotive segment in the field of light detection and ranging known as LIDAR. LIDAR is a key technology to enable autonomous driving in the future.

Already today, our infrared emitters and detectors are important parts of LIDAR. Let me just briefly explain this.

Infrared emitters and detectors help vehicles to generate data of their surroundings. The software and algorithm of LeddarTech convert these signals to information which are important to create precise images and to assist the driver.

This investment helps us to increase our position in the growth market of autonomous driving. Going forward, we will use similar investments and partnerships to further expand our strong technology base to profit from growth opportunities.

Just recently, for example, we took over the company LED Engin from Silicon Valley to strengthen the professional and industrial applications unit of SP. Via our start up in Cubata we recently acquired a stake in a horticulture company.

With that, we are further enhancing our position in the digital Ladies and gentlemen, let me conclude by giving you a brief summary. Through our hi-tech products we are able to grow in a profitable way.

We will use partnerships, equity investments and acquisition to expand our capabilities. Our access to the technology and to grow our customer base, as a provider of smart digital solutions we will provide new applications in autonomous driving, intelligent light control systems and for the digital industry.

OSRAM is successively on the way transforming itself from a supplier of components into a technology company delivering full systems. So for that, thank you very much for your attention and Ingo will give you now more details on our financial results.

Thank you, Olaf, good morning, good afternoon, thank you for joining the OSRAM earnings call for the third quarter of our fiscal year 2017 and I will now go through some of the key financials for the company in the quarter. Revenue, I am on Slide 6 now.

Revenue for the third quarter was EUR 1.056 billion, representing a comparable growth of 3.2%, relative to Q3 2016. The impact of foreign exchange accounted for approximately 80 basis points when compared to the prior-year quarter, portfolio effects amounted to 220 basis points.

When we consider the significant carve-out related to prebuying effects of last year's quarter of approximately EUR 47 million, growth underlying this quarter would have been closer to 8.5%. Opto continued its strong momentum followed by SP, LSS did not see market momentum pick up in the third quarter.

From a geographical perspective APAC continued to be our growth motor, growth in EMEA was slightly negative with 0.9% by and large driven by the tough comps given prior year's carve-out related to prebuying. In the Americas, market conditions continued to be challenging for LSS business.

When correcting for the prebuying effects, growth in EMEA would have been approximately 8.7%, 1.1% in the Americas and 14.4% in APAC. Moving on to profitability.

OSRAM's adjusted EBITDA margin for the quarter was solid at 16.4%, in line with our expectations and was EUR 174 million higher than Q3 fiscal year 2016 despite the prebuying effects, reflecting the strong growth in Opto and the margin resilience of SP. Compared to prior year's quarter profitability was lower by 60 basis points, driven by higher RD spends and lower profitability in LSS as a result of a lackluster volume growth.

We should also take into account that the carve-out related to prebuying had a positive effect on the profitability in the prior-year quarter estimated to be around 60 basis points. Corporate items were EUR 21 million lower sequentially.

We anticipate a somewhat higher run rate for the last quarter of fiscal year 2017 similar to the fourth quarter of fiscal year 2016. Special items this quarter amounted to EUR 27 million, the majority of which was related to the LSS segment, including measures to lower the cost of the organization and optimizing our manufacturing footprint.

For the full year we expect special items to be between EUR 75 million and EUR 85 million. Moving on to net income and earnings per share on Slide 8.

Diluted EPS from continued operations in the quarter reached EUR 0.65. This was EUR 0.10 lower than a year ago largely due to higher special items in the third quarter of this fiscal year and a difference in the effective tax rate where we had a one-time benefit last year related to the tax exemption of the sale of our FELCO share.

The company's corporate income tax rate of 29.6% in this fiscal year's third quarter was in line with our expectations. On July 10, we concluded our share buyback program as scheduled, against a total spend of EUR 399 million, we brought about 8.4 million shares representing approximately 8.1% of the total OSRAM share count.

Let us now move on and see how our segments performed in the third quarter of fiscal year 2017. Opto continued on a strong growth path and delivered 18% year-over-year comparable growth.

Please note that this business was not subject to prebuying effects in the same quarter a year ago. Growth for Opto continued to be driven by APAC as well as Europe.

Our infrared business powered further ahead driven by demand for mobile consumer devices. The SSL business improved momentum for into the third quarter.

Automotive and LED demand stayed strong fueled by good growth in Europe and APAC. Adjusted EBITDA came in with EUR 126 million translating into a growth in absolutely EBITDA earnings of close to 19% year-over-year, the 28.6% Opto's adjusted margin EBITDA margin was very much in line with the same period a year ago whilst the business continued to invest back into future growth.

Moving on to Specialty Lighting now on Slide 10. SP delivered EUR 563 million in revenue, representing a comparable growth of minus 2.3%.

When adjusting for the prebuying effect of last year's quarter growth would have been approximately 4% in the third quarter of fiscal year 2017. With the third quarter we entered the seasonally speaking typical low growth period of the year with demand for energy components and modules still robust but lower demand for traditional light sources in the weaker aftermarket business.

Adjusted EBITDA came in at EUR 73 million, very much in line with last year, despite a higher share of LED component in the revenue mix hence the comparably lower overall contribution from traditional light source technology given the prebuying effects in Q3 fiscal year 2016. SP's adjusted EBITDA margin profitability was 13%, close to last year 13.4%.

Looking ahead, we expect growth to increase in the fourth quarter of fiscal year 2017 supported also by easier comps related to the adverse prebuying effects in Q4 fiscal year 2016. In addition, we also foresee the typical seasonal summer patterns for this business to play out.

For the full fiscal year 2017 we expect SP to deliver an adjusted EBIDTA profitability of around 14%. Moving onto LSS.

I'm on Slide 11 now. Lighting Solutions and Systems revenue growth are flat when taking carve-out related prebuying effects in the prior-year quarter into consideration.

This segment generated revenues of EUR 253 million in the third quarter of fiscal year 2017, up sequentially. LS saw a better development of its revenue in the United States in the third quarter driven by an improved order book conversion.

Overall adjusted EBITDA was slightly negative with EUR 4 million for LSS improving sequentially but falling short of the performance in the same quarter a year ago. During our Q2 fiscal year 2017 call we announced our intention to take structural and cost measures in our LSS business in Traunreut, Germany as Olaf explained.

In the meantime, we were able to conclude an agreement with employee representatives. In the quarter we recorded approximately EUR 6 million in charges related to such agreements, with a bigger balance expected in the upcoming fourth quarter of fiscal year 2017.

This is fully in line with earlier communications. Given our LSS performance in Q3 fiscal year 2017 we no longer expect this business to breakeven for the full fiscal year as we do not anticipate that the current market environment is to develop more favorably on the short term.

Further actions are underway in LSS to adjust our cost structure to better align with near-term market growth expectations. Moving on to free cash flow in the quarter, on Slide 12.

Free cash flow came in at a positive EUR 39 million for the quarter despite an increased level of capital expenditures largely related to Opto. Working capital reduced compared to the same period a year ago, translating into better overall days outstanding.

We anticipate capital expenditures to increase further in Q4 2017 as we are nearing the production start in Kulim, Malaysia. Moving on to net liquidity on Slide 13.

We closed the quarter with a strong level of net liquidity of EUR 561 million, nearly unchanged to the opening balance of the quarter with increased levels of capital expenditures funded through operational cash flow in the quarter. I would like to point out that the acquisition of LED Engin and of the minority stake in LeddarTech will be accounted for in our fourth quarter of fiscal year 2017.

Let me now change perspective and look forward. Our guidance for the fiscal year 2017 remains unchanged.

We expect comparable revenue growth of fiscal year 2017 to be in the range of 7% to 9%. Adjusted EBITDA is expected to come in at a range between 16.5% and 17.5% for the year.

Cash flow is expected to be around the breakeven level and diluted EPS from continued operations is expected in the range of between EUR 2.70 to EUR 2.90. And with this, operator, we can now start the QA session.

I got two questions please. Firstly, can you tell us a bit more about LeddarTech, what the revenues are, whether you are going to be an active or a passive investor and what your plans are medium-term, whether it's an early investment, potentially followed by a bigger investment.

So some more color on that would be appreciated. And then secondly on Kulim, you are getting close to starting production now, can you talk maybe in non-numerical terms about how much the capacity is presold for 2018, how the mix is looking?

Just to help us understand how the business is going to develop in 2018 because I guess you must have quite a lot of the building blocks in place now commercially for the revenue ramp up next year.

Peter, it's Stefan speaking, maybe I take your question in regards of LeddarTech. LeddarTech, as Olaf had mentioned, is a software- and algorithmic-based company supporting the systems for LIDAR.

The forecast of the revenue in the next year is to go into a 3-digit billion number. As you know, this is, the LIDAR market is a very strong forecasted market because we will play an important role in the context of autonomous driving.

We feel LeddarTech as a strategic investment for technology. Our strategic rationale is twofold.

On the one hand side we see a strong development and a strong need for development of affordable LIDAR systems for the automotive sector and LeddarTech is basically an enabler to make very cost-optimized systems. And this is important for us for the OS business.

So being involved in the LeddarTech development gives us the opportunity to design our components, our infrared laser and laser diode components in an appropriate way to play basically an important role in the future of this reference design. The second opportunity which we see with the investment in LeddarTech is if you think about these LIDAR modules, at the end of the day you need somebody who has industrialization competencies to make these modules for the usage in automotive environments.

And I think with our core competencies in manufacturing of Opto electronics components for the automotive requirement in regards of volumes and qualities, we think that we can be an important player in the future in LIDAR. And these are basically the two strategic rationale, the two investment into Leddartechnology.

The current call, the final call enabled us only to participate in the range of 25%. How we continue with our investment we have to see it.

And that will be basically based on the development of LeddarTech overall.

Yes, and maybe let me just briefly add to that, we also do have [indiscernible] in LeddarTech, which I think is important and with that we also gain access to better understanding how that market and possibly also alternative technologies will develop. Peter let me address your second question maybe and Olaf can also chime in on Kulim.

So we are on schedule to start production in November of this year. And I would characterize the fiscal year 2018 of a year where we will slowly ramp up that capacity in Kulim.

And based on my understanding right now we will probably add at the end of that year roughly 5,000 wafer starts per year sorry, per week obviously. Given that it's going to be a ramp throughout the year, we will see basically how we have the mix.

Right now, the idea is that we have somewhat of a mix between our premium technology UX:3, and also, of course, the introduction of our new technology around sapphire. For both we have a lot of interest from customers, so we are quite confident that we are able to fill Kulim over the year.

But again it is a slow ramp, don't – please don't expect that the full capacity is available at day 1, that will be completely unrealistic. And the other caution I would also give with your models is that we will of course also take the opportunity some of – that for some of the chips that we now source externally that we then start to do that in-house so that we incrementalize our internal capacity utilization, which not necessary will always translate into revenue growth.

Maybe to add okay – Peter?

Yes. I was just going to follow up and say the 5,000 wafer starts per week by the end of the fiscal.

Can you give us an indication of what sort of capacity increase that is from what you have currently because I think you on about 2,000, 3,000 currently?

No, we are more than that right now. So it's not some – I think there is some ideas in the market that we all of a sudden would double or so our capacity or even more, that's not the case.

Right now our capacity is probably more in the neighborhood of 6K wafer starts per week if you take all the locations that we have across the globe. So this is adding maybe is – it is not doubling it, slightly less than that.

But again, it's a ramp throughout the year, and with the ramp there will be initial yields that will not look like the yields we currently have in Regensburg for instance. So I would be very careful to calculate that full 5,000 wafer starts per week for the full year, that would be a mistake.

And maybe to add on, we are, as I said, we are trying to speed up with Regensburg with that we are talking to the construction company going in a third shift that we are 6 to 8 weeks faster with the production and with the building and then with the production. So my concern is not the wafer start.

The capacity what is coming, my concern is that we do not have enough capacity.

Because I was looking at the grades that you got in Opto, I mean 18% in the third quarter, it was slower than Q2 when you had the prelaunch inventory build up for mobile phones. I mean 18% is still a very healthy number.

Were there any other unusual effects in the third quarter with prelaunch inventory build or is that more the sort of the run rate you are experiencing currently?

As you said it's really an excellent run rate, 18% is really excellent and compared to the whole industry. We don't see any slowdown, but again 18% is a high number and I would say I would be happy to have something between 12% and 18% whatever it is.

But today, as I said, we are fully loaded, we have really an issue to fulfill all the requirements from my clients and customers and I would be happy to have some of the capacity much earlier in the market but not before next year we will be able to come in a normal range.

A couple of questions, please. First one also addressing Kulim, we spoke about the wafer ramp up, can we also speak a little bit about the ramp up cost of the factory?

I mean at the moment [indiscernible] is penciling in 100 basis points margin decline next year for Opto. Conscious of the fact that you're not providing any detailed guidance it's probably.

But do you think that's going into the right direction? And then maybe I ask the next one after you answer this one.

Okay, I think it's a good question for Ingo but – I think penciling in 100 basis points is probably on the low side because, again, it's a very significant facility, it comes with a technology around sapphire that we don't have yet in any other of our locations. And like in any industrial ramp up in any other industry you will have in the beginning bottlenecks that you have to take out.

So you have under utilization, you will have material consumption that is higher than normal, you will have yields that will be less than the ones that are you are used to, and it will take a while. But it doesn't mean that we're not investing into sort of avoiding as much of that as we can but I think we should pencil in a little bit more than that honestly because I don't think that 100 basis points will be enough.

And how do you think about Phase 2 ramp because if I remember your original presentation on Kulim I think the Phase 2 should be decided this year. Are you on schedule with that?

We are absolutely on schedule on that but as I said so the original schedule is still in place we will do the modules. The next decision to go in module 2 will be earliest in the beginning of 2018.

So we make the ramp up. And as Ingo said we'll start in November, so I expect something in the beginning of 2018 to come up with module 2 or the decision and the discussion about module 2.

And then just coming back on the margin issue. I mean obviously everybody in our country talking about dollar weakness, I'm more curious about the positive impact you see from the weakness in the Malaysian ringgit, which I guess has an important influence on your cost base for Opto.

Do you see that as a potential tailwind or is it also over-compensated by the dollar weakness for Opto specifically?

No, as we had the advantage of the weakness of the dollar it was always an advantage for our project in Kulim. So we had tailwind, as you said, especially in construction costs and of course the whole running costs for the current factory to be honest, we have not only Kulim, we have 6,000 employees in Penang that's across Kulim and that's in fact an advantage for Opto today.

And maybe to add to that, Sven. We basically try to also run Opto on the basis of an exchange rate that the – where we don't sort of rely on the tailwind from foreign currency.

When we make the investment case for Kulim we measured on the basis of certain exchange rates and we continued to push the business in that direction to make sure that the profitability is such that you're not always depending on a weak Malaysian ringgit for instance. In the quarter itself the dollar only weakened towards the very end, that's also why we still had actually a positive currency impact if you compare to normal growth.

So from that perspective that's going to few more on the revenue line probably going into the next quarter and possibly next year depends little bit on where currency markets will go. The ringgit is typically very much linked to oil from a – there's a high correlation between the Malaysian ringgit and the oil price, so we have to also see how that is going to develop.

But again we're not sort of banking on the ringgit to improve or reduce profitability in that business.

And then I was just wondering in terms of what you said on LSS that the breakeven is obviously not possible this year. So should we assume that to be the target for next year then?

We're currently going through our budget process. A very important aspect of that assessment is also how some of our key markets are going to develop.

The – as I said in the prepared remarks, the North American non-residential construction market is still in sort of sideways, in some cases even going down. We don't see any signals yet that that would significantly improve over the next few quarters.

So we are still evaluating that. At the same time we also, as I said in the prepared remarks, revisiting the cost structure of that business to see what we can do to faster align the cost structure with the market outlook.

So for that we're still going through that process so I don't want to take ahead what we will say. We will update you on that specific question when we result on – when we present our results on Q4, the full year, and Olaf and myself will give first guidance for fiscal year 2018.

The last question I have is just on the announcement you had mid-July on the potential joint venture with Continental, if you could give us some flavor of that and whether you see a chance of these talks gathering some speed again or what's your color on that?

As we announced in July with – we are in good discussion with Conti about a potential joint venture for automotive electric lighting, but they are really no new news. I can tell you today its vacation time in Nidavexin and we will come up with the next phase after, to be honest, after the vacation time.

I'm going on vacation tomorrow and so we're coming up with – I'm sure that we're coming up with some news in some weeks.

I will have three. The first one is just maybe to understand a little bit better why you haven't gone the full way in your share buyback considering that you had, I think, initially planned EUR 0.5 billion potentially and you've done a little bit less than that?

That's something for Ingo.

So we always said that when we communicated the share buyback from the get-go that we would spend up to EUR 500 million, we never said we would spend the full amount, and we felt at some point in time when we looked at our priorities from a capital deployment perspective that these priorities were not actually not so much with buying back our own shares but rather investing into the business. And therefore we prioritized that we do focus more on building the business with our capital investment programs that we have predominantly in Opto.

Also given the new opportunities we have there on the infrared side of the house and also by intensifying our lookout for MA targets.

The second question I had was around the joint venture or kind of the partnership between Cree and Sanan. I was wondering and of course I appreciate it's not for the next quarter but a bit more medium-term or long-term, how you were seeing this kind of partnership and also because Sanan in their latest report said that they have improved their certification process for their auto-related business in China.

I cannot tell so much about the Cree and Sanan because I have the same information like you from the press. That means it looks like that they are looking for a partnership for a joint venture in Hong Kong.

It looks like that – and what I know is from the past that Sanan and Cree vice versa bought chips, so in the past Cree already bought chips from Sanan. So I have really no detailed information about that.

The second question about automotive, to be honest there everybody is talking that they would like to go in automotives business. And what I can see is you need to deliver the automotive industry, you have a long time for qualification, you have to be certified.

And as you know, Sanan, they don't have the licensees and they are not certified. So the wish to go in the automotive industry is one thing, to be in the automotive industry is a different story.

So I'm – I don't see them any – in any – with any customer.

And just a last question and that actually arose from lunch we had with lot of the suppliers, a discussion we had with them, I was wondering when you are kind of present on the model for car that you supply chips for that is going to be included in a broader system, for example a head lamp, who is actually choosing the chip supplier or the chip manufacturer, is that the actual OEM or is it the OEM supplier which is kind of putting together the biggest module?

Very interesting question. Overall we see basically a trend with more complexity in the light system and especially what you mentioned front light systems, it's a thing about adaptive front light with more than one LED pixel that we are involved more and more in the topology of the system architecture of the car itself, and therefore we see more and more development activities directly with the car manufacturer.

And basically when we have developed a solution together with a car manufacturer we see what we call a direct buy. But basically the set maker who is providing the light module or light system gets basically a clear direction that we have to be designed into this front light.

So with increasing complexity we see more and more direct approach from the OEMs to us as, let's say, an electronic, an Opto electronic provider for modern front light systems.

And maybe to add on that, one example is the Volkswagen company. We are not only preferred supplier; we are one of 52 suppliers worldwide.

They are in a very close research and development cooperation, that means we have an early involvement about the technology vice versa to see what they would like to develop and we are coming up with ideas as well. And we are the only lighting company worldwide from the 52 suppliers and there is no set maker inside, so that means the OSRAM itself is now very close to the OEMs.

Just if I can make a follow-up on that in terms of your relationship with the then the OEM supplies as well going forward, because due to some extent I think also when you get more integrated into the system and also the full-on LIDAR module and so on you are kind of it looks like getting in competition with some of the guys which used to be your customers or who are still your customers on maybe other type of products, how do you manage this relationship?

To be honest, of course, but that in these days is really a normal situation that they call it that in – with some customers you are in competition, in some others you are in cooperation, so that's one of the reason that is the name competition, that means of course with some clients OEMs they have a clearly view that we have to work together the set maker and the OSRAM and in some other they would like to get the full system. So this is not really a conflict and you know in some other industry, it's a normal case.

Think about SAP, IBM, Microsoft, sometimes they are competitors and sometimes they have to cooperate. So it's something normal I would say.

Congratulations on the quarter. A couple of questions.

First, going back to LSS if we could, Ingo, your comments around North American nonresidential still being week. There are some major lighting companies that said things got better.

Is that just sort of temporal channel shifts in mix issues? I know you have some insights given your driver business, but what do you – what are you attributing the weakness to because after several quarters of weakness that seems like at least some like Acuity are saying things might be better.

Yes, so we are, I mean we are at like Acuity we are part of NEMA, the North American Electronic Manufacturing Association. So we're using typically those data because they cover 60% to 65% of the market.

So I think that's always a good source. And you see that overall we see a very cool nonresidential construction market, which is typically our key market because we are not playing in the residential market there in the United States.

We see that particularly the small-and medium-sized projects with architects, et cetera, are not coming back the way they were about a year ago, and a lot of that seems to be still related to uncertainty with respect to tax reform and people still have a chance to push out these investments because the demand doesn't seem to be that strong. There are a number of bigger projects around that's for sure.

I think Philips lighting also said that the same, we see that as well. But also it takes awhile before you are actually commissioning larger projects.

And larger projects typically mean that first the refurbishment and the construction has to go ahead and then as a lighting or a ballast supplier you are at the very end of the supply chain. So there is a significant lead time between the order intake and the actual sell through that we have and that's simply what we're still seeing in the United States.

Okay. That's helpful.

Thank you. On the infrared chip side, can you just give us an update of what you're seeing in terms of design wins and expectations for either refreshed products in the handset side or new applications, if you could just give us an update there on the IR side.

So we have still a very strong demand especially as we talk about that one of the mobile phone producer had a huge demand for infrared as well. The demand for the next generation for mobile phones are coming up with the same iris scan and gesture recognition.

So we won some design wins and I'm quite confident that we will come up with additional clients and customers for this business field as well.

Yes, maybe add to that, I mean Olaf was mentioning the investments we've authorized in Regensburg here in Germany that – the big part of that investment is actually geared towards expanding our infrared capacity. So the demand at this point in time that we have for infrared product is bigger than the capacity that we have in house.

So we will be for a while until that new facility goes live, we will be capacity constrained.

Okay. Thank you.

My last question more strategy around moving into partnerships and joint ventures, how exactly do you decide if you want to just buy a little small company like you did with I think LED Engin versus a minority interest versus a JV. Just talk to us, if you would, around your process around entering into these relationships with third parties and build versus buy and things like that.

I will – Stefan and me will answer that. I think in general we are looking really for good opportunities.

If we have a chance to buy smaller technology IPs we will do this like we did it with LED Engin and the small company in Silicon Valley, very interesting for our peer segment. In some other like the LeddarTech it is partially investment and I think we have really an open view about all possibilities.

And maybe, Stefan, you can talk a little bit about –

I think what we do basically as a rationale when we think about this cooperation, joint ventures is we have basically a map with two scales,one is how can we directly support our business and what is helping to build up competencies for the future. And what Olaf just mentioned, LED Engin is something which immediately will support our business at Clay Paky in the entertainment lighting because there we see a very fast transformation from the classic light sources to LED-based light sources, and LED Engin has the system architecture and the qualities for LED and electronic to support basically this entertainment business immediately.

Second investment into LeddarTech is more on the other scale where we say, okay, what are future competencies which we need. And as I said before, there are two strategic options; one is basically to support our component development by early knowing of the requirements in these new systems for laser diodes, for instance; the second is to see which kind of LIDAR modules will be developed and how can we basically bring in our competencies in the industrialization of these components and modules for automotive needs.

So the second one is more competence build up rationale and the first one, LED Engin, is more a business supporting rationale. And that is how we basically discuss each and every object and issue which is coming up on our MA target list.

And as we said, it's over the question of time, speed, we want to be fast. So if we can develop something in-house we will do it in-house.

Is it faster? If we can buy something in the market then we go maybe this way, and maybe Stefan, could you give us an update about VCSELs light that you talk a little bit about that.

It's an interesting technology as well what we have on our radar screen.

[indiscernible] speak to what we, I think, also discussed this around in the last quarter really. We are currently doing an analysis of what is basically the market need.

VCSEL is one option for a light source next to laser diodes, infrared laser diodes and infrared LEDs itself. And it's basically the question of the application in regard of the mission angle and for some of the applications VCSEL obviously appears as the appropriate solution and we are currently doing a market analysis to see how big is the market.

Technology wise, it is not a big issue to implement basically a VCSEL production into our current setting in Regensburg. However, it is a business decision if at the end of the day we see enough volume.

And then basically one of the next MA targets could be as business supporting also an activity in VCSELs. But that's basically the rationale how we decide this MA targets on the one hand side competence build up, and on the other hand market potential and support market and sales build up for these new market segments.

Thanks for taking my questions. I also have three and they are all on Kulim, so perhaps I could go one at a time as well.

Firstly, could you perhaps help us with what the equivalent as far as wafer starts per week is of the external chip sourcing that you will internalize. Basically, how much of the 5,000 in Phase 1 will that represent and will it all be internalized in Phase 1 or will some of it spill into Phase 2?

Sorry. Can you clarify your question because I'm not sure I completely understood it?

Sure.

And I want to make sure you get the right answer.

Perhaps, I'll try and re-explain it. You mentioned that we shouldn't over anticipate volume growth in Opto from Kulim next year because some of it in terms of revenues will be you internalizing sourcing that is already done externally.

And I was thinking how much of the 5,000 of Kulim Phase 1 is effectively that 1 out of the 5, or 3 out of the 5 that represents the internalized volume?

Yes, now I understand the question. Thank you for the clarification.

It's very difficult to explain it in the type of wafers or size because depending on the product as you sell you have more chips in one than the other. So from that perspective I wouldn't really go down that path.

At this point in time if you look at the current revenue base for our SSL business, if you look at revenues, just revenue, we probably saw it in any given point in time anywhere between 15% to 25%. Externally, that can be dies, that can be chips.

So it's also different on the application and where we might have capacity shortages, and we go outside, et cetera. So anywhere between 15% to 25% of the current turnover of SSL would be subject to possibly internalization.

I don't think it's going to internalize next year, but that's a kind of magnitude that you have to bear in mind.

And secondly on thinking about the premium GI mix not on the whole of Kulim where you have been clear, but just the first phase, should I think of it is all being GI given auto certification and infrared chemistry isn't going to be there?

No, it takes a little bit time as I said in the last call or maybe the call before. We are running, we are doing our best to get the automotive qualification as soon as possible, but that takes usually takes time.

And I can start of course in November and not before. So we have to run it up and then we come up with the qualification for the automotive, and that will be not – that can't, it can be only in 2018 and mid of 2018, end of 2018 and not in the beginning of 2018 if we have the ramp up in November 2017.

But the second question is with the mix, the mix is going up more and more to premium. As I said we have huge demand on premium chip and it is clear from the margin point of view we do our best to sell premium chip instead of a GI chip.

And today what we see is that the demand is high in premium and we will bring, as soon as possible, more premium chips to Kulim.

Just to make sure that we don't have a misunderstanding here, so it's not just automotive that is premium, okay. There are also other products that we sell today that will be considered premium or UX:3 based if you like.

So from that perspective we will produce right from the day 1 in November premium related products out of Kulim, but not for the automotive industry. Rather what we are doing industry I think we explained it at another occasion, is we're going to try to shift some chip manufacturing from Penang also in Malaysia into Kulim.

And therefore what we do is we create more capacity for automotives in Penang because Penang is automotive certified and hence premium manufacturing will move from Penang to Kulim. We produce therefore premium in Kulim and we open up more capacity for automotives in Penang.

I know it's complicated, but that's the way it is. And therefore, as Olaf said, we believe that the mix will be more favorable in Kulim already in year 1 relative to what we discussed 18 months or so ago when we had Capital Markets Day.

That is very helpful, thank you. And lastly, if I can, just conceptually could do with some help on thinking about the average selling price difference between GI and premium.

And one of the reason I'm asking is because the price mix of Opto has been a pretty steady number over the last few years. I wonder whether that's going to change materially next year as we end up with a greater degree of actually GI mix, I am just wondering how that could move.

I think that's a question that is – if I were to answer this I would be probably misleading you, because it changes a lot. Even within premium we have very big price differences, you have big price differences even in general lighting, because in general illumination there are market segments that are actually professional grade based on specifications and there are more consumer type of grades, which have very little specifications therefore also different ASPs, there are different packaging, et cetera.

So I'm not – I don't want to do that just not because I don't want to tell you what they are, but just because if I give you a number now you would just, I think, making a mistake in the model. So I will refrain from doing so.

I have got 2 or 3 questions, 2 on Opto. Firstly, Ingo you have been so kind and gave us an indication on the margin for next year when it comes to consensus and said that 100 basis points maybe a bit on the low side and at the same time you obviously – sorry to follow up on this – but you indicated that we shouldn't overshoot on the revenue mix.

However, just doing the math and even appreciating that you will have some internalization effects from the LEDs that you are so far sourcing out of Taiwan. Assuming that, I don't know, on average 30% or 40% increase on the volume side, how do you square that up with the consensus expectations of 14% growth for Opto next year, just some additional color would obviously be helpful.

And then secondly, I have an follow-up on Opto really and we did see one or two capacity announcements out of China from yes, Tier 1 players that are not Sanan.Just wondering how you – whether you believe those are replacement of these older tools from the big boom time in 2009-2010, whether you thought the new capacity is competitive as far as your strategic marketing is really concerned, and would that kind of influence your CapEx decision for Phase 2 in Kulim, which is due, as Olaf said, early next year.

Let me start with the first question and then Olaf will talk about the competition, et cetera. So I, frankly speaking, I didn't look into the consensus for next year for Opto.

As I said, we are still going through the budgeting process. It would appear to me though that the number that you mentioned would be indeed probably underestimating what kind of growth trajectory we should see for Opto next year, how much more than that I'm not going to disclose at this point in time, because again we are going through the budgeting process, but that seems to be more on the lower end of the spectrum that I would expect.

And to Uwe, coming to the second question. In fact what happened in the market is what we really proposed on our road shows and in so many meetings.

I said we will have a balanced situation in 2017 and we are coming really in a position that maybe we have an allocation latest in 2019, because the demand of LED is high and some of the old machines were going out of the market, that's what I really said in so many meetings in London and Frankfurt and New York. And if you take a look maybe to 1 of the press what came up 2 weeks ago from DigiTimes from China and they wrote that they were – that China – in China they will retire 1,800 old MOCVDs.

Again, the whole market has 3,000 and 1,800 are going out and they're in 2 and 4 inches. So that means over 60% percent of the whole capacity is moving out because it doesn't make sense to produce in a 2-inch machine.

So on the other hand, I said in the market as well, we are looking forward that new investments will happen. It can't be that it will be only Kulim.

And 2 Chinese LED producers announced that they will implement between 15,000 and 20,000 wafers starts per week capacity in the next years. So we have exact what we proposed, old machines are going out and new capacity is coming in.

That is very clear. And just on your Phase 2 potential capacity announcement that you indicated would be early next year, is it also, let me say from a negative perspective, can you completely rule out the investment into infrared in Malaysia now that demand turns out so much stronger than expected or can it be only handled in within Regensburg expansion from today's perspective?

From today's perspective it can't be only in Regensburg, it is my wish in the mid-term to move to Malaysia, could be Penang, could be Kulim. Again we have a huge factory in Penang as well.

It is my wish to have infrared in Malaysia. But from today's perspective, and I'm talking then really for the next 3 years, it is impossible.

Should be because of qualifications issues?

Yes, yes I don't – I do not have the qualification, that is my point, I do not. It's a special technology, you need special people for that.

And again it would be my wish but it's impossible. So that's one of the reasons we invested in Regensburg.

So that's the main reason why we invested in Regensburg. So we doubled the size of the Regensburg plant.

And if you would come up you will see that.

Alok Katre from Soc Gen. Thanks for taking my question.

Couple of things. One, I think you've been mentioning that actually you're falling short of capacity and therefore you had to sort of thinking about pulling forward Regensburg, trying to get Kulim up to speed as quickly as possible, reshuffling production between Kulim and Penang, et cetera.

My question then especially on infrared side is what is the risk that some of your competitors can actually take advantage of the fact that you are running short of capacity and does that then, let's say, is that a bit more of a structural sort of constraint from the future growth perspective or do you think that there is actually not many people who can do these products that you do and therefore customers will wait. That was question 1.

Second, just on LIDAR interesting I mean chart that you flag off right on – at the beginning of your presentation with 35% CAGR in revenues. Could just give us a sense of what market size you're talking about and how we should think about your position here alongside the competitive dynamics, I mean are we talking a few percentage point extra boost to the growth at the group level over the next 2 to 5 years, is that sort of what's the sense over there?

And the last one, if I can, just on the 2020 targets and the discussion around Kulim and expectations et cetera. Is it fair for us to think that given the growth profile you've seen at Opto higher mix of infrared than [indiscernible] et cetera versus when you set your 515 targets 18 months ago.

That of course, it seems you are probably then ahead of the plan and in that case what can go wrong from here to not allow you to, let's say, exceed your 515 targets.

Okay, Alok a lot of questions. Let me start with question 1 with infrared.

Again usually I don't like it so much if you are really alone in the market because always if you cannot deliver and you are alone, the whole competition sees that you can make good margins and try to step in. It's always better to have a balance situation.

The advantage of OSRAM is that we have the IP rights. So it is not possible for someone around the corner to start with the infrared.

Two reasons, again you need special technology and knowledge, and second one, you need the IP rights. And for this reason I'm not the afraid that somebody unknown would step in, in the infrared chip business.

But nevertheless, there are some of the market they could be but again it takes time, you need capacity, you need machines to ramp up and that's one of the reason I speed up in Regensburg with the capacity and speed up with the construction to come as early as possible in the market. And we're talking about really every single week is helpful.

And the second.

So is it fair to then say that the infrared market as a whole is far, far more tighter than, let's say, any other part of the LED chip market? Meaning basically no one in the market has extra capacity and everyone is running flat out?

Now, capacity is one thing but you need – nobody has infrared capacity. That's right.

That is not in the market. The second question was about the LIDAR market.

And the LIDAR market, as Ingo said, is till 2020 around EUR 1 billion. But the second number is that in 2026 the market is expected by EUR 2.5 billion.

So we have a CAGR by around 30% of year-over-year and that means it is a growing market. And as Ingo and Stefan said for the autonomous driving you need these technology and the autonomous driving would come up.

You see that we get more and more driver assistance systems. So every new car has additional features for more driver assistance systems.

The new Audi A8 has now Level 3 and you will see that the next generation will give more extras on the way on the autonomous driving. So that's a great market, attractive market, and that's one of the reasons we step in.

So the third question was something about 515. That's for you, Ingo.

Honestly, can you repeat the question?

Yes, sorry for that. So just thinking, when you set the 515 targets and since then what we've seen at Opto in terms of the growth profile this whole infrared, auto LED and the LIDAR, et cetera, et cetera.

So clearly it would seem that you are well ahead of plan in terms of 515 target. So what I was trying to think was what can go wrong from here for you that will allow you to – or let's say that would make you to not deliver on the 515 but also not exceed the 515 target.

It seems you are very well placed to even exceed those targets. Just going around the discussions that we've had, Kulim capacities, sales expectations, et cetera?

Alok, thanks for your confidence in the company. Thank you.

It is certainly true that we are on a good run with Opto and the numbers show that. And I think the numbers also, particularly with Opto, show that the investment case we represented some time ago was the right decision to make.

We are also quite happy with the development that we have within SP. At this point in time however I think it's also fair to say that the performance in LSS this year is a bit of a setback for us.

As we clearly said at the beginning of the year that we would target for a breakeven situation, which we will not accomplish as I said in my prepared remarks. So from that perspective we are focusing right now on our budgets for next year and I think we will get the next outlook for fiscal year 2018 then around the November or so timeframe.

And I think it's very premature to now even talk about 515 other than in the terms we indicated, it's still the targets for the company, it hasn't changed. And I think we're sort of only in year two or so [indiscernible] or even year 1 after we announced it.

So it's a bit early to change that.

Actually, Olaf answered most of my questions already. Just wanted to clarify when you say LSS breakeven will not be achieved this year, are we talking adjusted EBITDA, EBITA, or what is the basis for your statement here?

We always said that the breakeven target was for on the basis of EBITA not –

EBITA, okay.

EBITA. So that will definitely not be accomplished this year.

I also don't believe, just to be clear, that there will be breakeven this year on an EBITDA level either.

Okay, not in EBITDA. And then just on Wuxi expansion, can you quantify that, how much should we be looking at that.

Is it a big addition or a small?

Well, I think Olaf said earlier on that we are basically building the second wing of that facility, so it was always designed to have two wings when we built it originally in Wuxi. We are now building the second wing which basically means that we would be doubling the existing back-end capacity that we have there and we will basically do a ground breaking in...

Two weeks.

In 2 weeks or so. But it will take certainly more longer than just the fiscal year 2018 to become live.

So I don't expect any major impact other than CapEx and maybe some ramp-up costs in fiscal year 2018.

And maybe to add on that as we ramped up Wuxi 1 you didn't see anything in our results, so they made it perfect and the same team as they ramped up Wuxi 1 will ramp up Wuxi 2. But to make it clear, we are absolutely full out of capacity for backend care, I have the same issue as I have in infrared, I have my issues with backend capacity.

Okay. How much CapEx are you planning for this?

I think we will give guidance to that not now, but in the next fiscal year, so early November, so stay tuned and you will get somewhat more guidance.

But don't be so afraid, it's easy, backend is easy, it's not frontend. Backend is easy.

And we could,

One follow-up question on the VCSEL regarding LeddarTech, so I was just wondering is this technology agnostic regarding whether you use for illumination conventional laser or VCSEL?

No, that's – what LeddarTech is basically developing is reference designs for certain applications. One for instance, for long range which is then supported by a scanning system based on a mirror, but also so-called flash LIDAR, where you illuminate the entire scene with one LIDAR source or with one infrared source.

And as I said before, it depends somehow on the application and LeddarTech is not only looking into automotive applications they are looking also in an industrial applications. And what I said before is also valid for LeddarTech, depends on the emission angle which you have to achieve in regard of the application needs and then it could also be a solution or an array of VCSELs.

Then I had a question on specialty lighting margin, the 14% guidance for the year. I was just wondering, can you quantify the negative impact this year, coming from the higher share of inter-segments sales you have with Opto, and so – but I don't know was 50 basis points, 100 basis points which was basically offset by other factors.

I think you can basically derive the inter-segment sales by the disclosure we do on our CIE segments in the disclosure. I think from there you can fairly – you could see what sort of differences are between last year and this year.

Also we said before that overall within SP approximately 3% lands on the bottom line from the transfer price they get from Opto with the residual being an Opto and then that also hasn't changed, so I guess when you do the math in that you can probably get to a result.

Okay, I'll do that. Then the last one is on LSS again and especially regarding the systems business, LED converters and so on.

I mean, in this business I remember you achieved the turnaround in the second half of 2015, so given the issues also regarding Luminaires' customers in the U.S., so I was just wondering, what's the earning situation right now of the systems business within LSS?

It's getting better from 2015 to 2016 and it's under pressure now, as Ingo said. The main customer is QT and Eton in U.S.

for example, so we have a high market share in the United States, we are clearly a market leader and both customers had decline in sales and this is one of the reasons that we have a weaker profit situation than the year before. We don't have a structural issue so it's really it's the market, it's not a structural issue.

Thank you for allowing the follow up. Just slightly more mundane question, but I was wondering if you could give us a bit of color around the gross margin compression you have year-on-year is 200 basis points and how we should think about gross margin dynamics from here?

Okay. A big part of that compression year-over-year is related to special items.

So we had a higher number of special items that went through the gross margin or cost of goods sold if you like than last year, it's approximately 130 basis points of the difference. So the compression is 60 basis points for the company overall and it's largely due to, I would say, two reasons; one is there was a significant pre-buying effect for SP business last year and that pre-buying was largely because of traditional light sources which have higher profit margins, as you know, Lucie, so that's one reason why it was lower.

And the other reason is what we just talked about at LSS we also saw a compression in our gross margin in our LSS business particularly in the United States. There is a lot of competition right now in the traditional ballast business, we see significant declines and we also see some people coming in with pricing that is very challenging.

So from that perspective that was the reason.

Yes, thank you very much for your question. And sorry for prereleasing the figures yesterday evening.

If you have further questions please contact us and we wish you kind of a nice summer time. Thank you.