Nancy Goossens - Former VP, IR Dominique Leroy - CEO, President and Director Bart Van Den Meersche - Chief Enterprise Market Officer Sandrine Dufour - CFO Daniel Kurgan - CEO, Belgacom International Carrier Services

Daniel Morris - Barclays PLC Vikram Karnany - UBS Investment Bank Ruben Devos - KBC Securities Emmanuel Carlier - ING Groep Ulrich Rathe - Jefferies LLC Paul Sidney - Crédit Suisse AG Stephane Beyazian - Raymond James Euro Equities Marc Hesselink - ABN AMRO Bank

Thank you. So yes, good afternoon, ladies and gentlemen and thank you all for calling in.

I take it that everybody has well received our second quarter result release and presentation. So as per usual, we will not run you through all the details, as we trust you have seen the published numbers by now.

So that leaves for me just to quickly introduce to you the people around the table here. So we have Dominique Leroy, the CEO; Sandrine Dufour, the CFO.

We have Bart Meersche, the Chief of the Enterprise Segment; Geert Standaert, the CTO; Daniel Kurgan, CEO of BICS; and Dirk Lybaert, our Chief Corporate Affairs Officer. They will be all be happy to take your questions in a moment.

But before getting to that, just the introduction of CEO, Dominique.

Yes. Good afternoon, everyone and welcome to our conference call.

I will just quickly run you through the main takeaways from our release this morning. I mean, first of all, I think we can all be proud of our ability to continue to grow the Proximus customer base and this in spite of a more competitive Belgian market.

Specifically, the Proximus All-in offers continue to prove successful with now 255,000 customers having subscribed to either Tuttimus or Bizz All-In. We close those who were very good quarter in particular for mobile postpaid, growing in total with 60,000 cards in the second quarter.

Alongside the commercial success of the Proximus brand, we also gained good traction on our Scarlet brand. Scarlet increasingly proved its solid competitive position in the low end of the market and successfully acquired customers that are more price-sensitive.

Our reported net customer growth was also supported by improved churn levels for fixed and mobile, something we can be proud of as it proves that the many initiatives we have taken to enhance customer experience has indeed turned Proximus into a more robust company. Also, our Enterprise segment firmly grew its mobile customer base and although we faced fierce competition in the Enterprise market, we sustained our overall solid position.

The second quarter revenue showed, however, some impact from the volatility in ICT product base which came on top of the ongoing decline in legacy voice services. While our domestic direct margin is facing some pressure due to a changing product mix and due to roaming regulation, we nevertheless achieved a 1.1% growth in our Domestic EBITDA.

This is thanks to lower domestic costs decreasing by 3.6% in the second quarter as part of our ongoing company-wide cost-reduction plan. The Domestic EBITDA achievements were partly offset by BICS bringing the total group EBITDA growth to 0.4% for the second quarter.

With these second quarter results, we remain broadly in line with our expectations and hence, we reiterate our full year guidance, meaning a slight growth in group EBITDA and a nearly stable Domestic revenue. With the final rollout now well started, we expect our Fiber investments to go up over the second half of this year and hence, we also reiterate our CapEx outlook for 2017 of around €1 billion.

Going forward, we will maintain our focus on improving customer experience. In this view, we have strengthened our entertainment offer with the renewal of the Jupiler Pro League football rights and we have added Netflix as a TV option in Tuttimus.

Also next week, we will launch our More4More data offers, both for mobile-only and convergent bundles. With these competitive offers, our mobile customers will be able to enjoy more mobile data, both in Belgium and abroad.

As a final point, as you know, the consultation on the BIPT's market analysis has now started and currently proposes to deepen the cable regulation and extending it to fiber network. Several modalities, however, still need to be defined.

Over the coming months, we will continue to strongly advocate for a nonregulated, open and future-proof fiber network based on flexible bitstream access. We will enable a competitive gigabit markets in Belgium.

We want to guarantee fiber access to Belgium households in the quickest and most cost-effective way, taking into account the specific Belgian context. We're also open, as I already said, to co-investments in fiber infrastructure, but this needs to be under conditions that are acceptable both from a technical and financial point of view.

With this, I have covered my introduction, but I am sure you will probably have more questions, so let's turn to this now.

I've got two, please. The first is directly following up on the comments you just made around the telecom regulation and fiber investment plans.

You mentioned that you're specifically looking for the right technical and financial conditions to do cofinancing. And I just wondered, have you looked to the French model which is essentially a kind of small margin on the build and then a shared-cost program.

Would that be a kind of acceptable solution in your view? Or are you looking for more in terms of the financial?

So any clarity you can provide on that. The second question I just have is on.

Roam-Like-At-Home. We've obviously had a little bit of time for that to start impacting, I guess, in the last few weeks.

Can you just update us on the kind of elasticity benefits you're seeing versus the headline reductions you've already guided us to?

Yes. On the co-investment first.

What I think is important to highlight is that we're open to co-investment which is also some different than pure co-financing. One of the proposals -- one of the elements where you see in France is that some operators are doing some parts of the cities and others other part of the city and then they make an agreement between themselves for open bitstream access.

This is potentially something we could consider, but therefore, you need someone that is ready to invest in fiber next to us in Belgium which today is not the case. On the Roam-Like-At-Home, I think we -- it's difficult to say, because it's a very recent figures already because it only went into the market from 15th of June.

What we definitely see is that the consumer are using way more there from abroad and we see some increase on our residential part for mobile data which can go up to 400%, 500%. It's less the case on the Enterprise business, where people have always used more their roaming abroad.

But we also see elasticity on the roaming in, knowing that of course, in the holiday period, we will probably have a higher elasticity in roaming out, where Belgian takes their holidays abroad and probably less elasticity in roaming in as Belgium is not so much a holiday destination, but way more a destination where people come for work during other months of the year. So a bit early to give more elements than those for the time being.

First question in terms of the higher income tax legal prepayment that you highlight that is impacting your cash flow. Is it more of a phasing issue?

Or there is an incremental burden on your free cash flow? And if you could quantify and what that additional burden would be in terms of this payment out.

And secondly, when I look at the fixed line overall in terms of enterprise side, your voice revenues are kind of under pressure and you've highlight that it's kind of ongoing rationalization that you're doing. Again, is it more of a phasing issue?

Or is it more structural that we should expect ARPU pressure and revenues therefore coming under because of what you are seeing as rationalization?

Okay. So on your question regarding the higher cash income tax prepayment.

So I think this year, there is a new framework whereby we have to pay 59% of all our tax due for the year by 15th of April, while last year, it was only 9%. So that is a change.

So is it just timing or does have an impact? For the full year, it's something that we had fully anticipated, so it certainly has a timing impact as you compare our first half free cash flow compared to last year.

Now bear in mind that our level of cash taxes that in 2016 were quite low. And that was to do as well to the fact that taxable days was lower in 2016.

And compared this year, we have less notional interest deduction, there are less other deductible. So we had anticipated that our cash taxes would increase in 2017 in our full year forecasted free cash flow.

And we still have in mind that we're able to cover our dividend with our cash flow despite the fact that our CapEx are growing.

And as far as the fixed voice erosion enterprise, yes, it is structural in the sense that we have seen this on the past quarters. It's been to rationalization of last, it's moving to more -- moved to voice over IP and it's linked also to some competitive pressure.

And so -- but if you see, it's fairly stable quarter-over quarter. The only thing that is different in this quarter is that in this quarter now, we have more, what we call equivalent -- less equivalent business days because the full Easter holidays were in the second quarter for May and April.

And so that plays in the usage of fixed voice for enterprise.

That's helpful. Can I just follow up in terms of the tax?

A separate topic. What is then your effective tax rate in 2017?

Yesterday, there was this announcement in terms of potential for the corporate taxes going down. What are your early thoughts on this particular topic if the government goes ahead?

What do you -- would plan to do? Would it be more kind of using the flexibility to accelerate your fiber rollout plan?

Well, the first elements, we know on the measures on the corporate tax reform are obviously good news for companies which are paying most of their taxes in Belgium, like Proximus. So before any decision, you know that are ETR for the year was trending closer to 30%.

And that's, as I said, just before because of a lower benefit of some deduction, precisely the notional interest tax reduction is yielding less than in the past because interest rates are very low. Now what has been announced is that if this is passed that tax rate, corporate tax rate, nominal corporate tax rates will decrease from 33.99% to first step in 2018 around 29.5%.

And potentially further down to 25% by 2020. Of course, there are series of compensating measures that will be taken as well to finance, this decrease of corporate tax rate.

Our first and preliminary assumptions and analysis of the impact that these measures could have on our ETR is that we would expect to have an ETR in the range of 27.5% to 28% as of 2018 which is certainly a decrease compared to the higher corporate tax rate that we were having. It may have an impact as well as of 2017 because of deferred tax rate but we're still calculating that.

I think it's highly premature to say that it has any impact on our operating plan in terms of deciding to accelerate investment or the like actually to be shared with you. It's not something we've been focusing on since yesterday.

The other element I'd like to mention is that there's an additional aspect which is linked to cash prepayment. I've just explained what's going on in 2017 with the higher prepayments up to 59%.

What's in the tax currently is that this prepayment could go up to 70% as of 2018 which mean that from a cash point of view, in 2018, we may have slightly negative impact. But all in all, ETR longer term, that's certainly -- good trends.

Two questions, please. First one on Scarlet.

So the press release indicated Scarlet's billing to customers in the low-cost markets, a message you've seen in Q1 as well. It all starts to appear in your triple-play ARPU as Scarlet customers increase in the mix.

So I was wondering whether you could give any qualitative comments on whether Scarlet sees most net that's coming from the existing subscriber base of Proximus or from your competitors? Whether it has been more successful in the south, in the north?

Whether you've seen more of a general EBITDA in the fixed market towards low-price bundles? So any comments on this would be helpful.

Second question, you've added 44,000 postpaid subscribers in residential which is again a rather solid performance. So could you give an indication whether the majority of the net adds intake can be explained due to the efforts you've done around prepaid registration and thus acceleration of the migration towards postpaid?

Or is it rather the impact of Tuttimus and then the conversion discounts which supposed to be -- you have a strong cross-selling incentive, let's say, to mobile?

Concerning Scarlet, it's true that Scarlet is working well. And that of course, when you sell a triple-play on Scarlet at €39 versus a triple-play on Proximus is around €65, you wait for longer term a small effect on the ARPU.

What is important to say is that we don't see more transfer from Proximus to Scarlet. I think that is relatively low and it's always been low.

And to be honest, it's also part of our strategy to try to target very different customer profile with Scarlet and Proximus, so we don't see transfer or major degradation of customer going from Proximus to Scarlet. What you see today is with all the new players on the market, we see more customers sensitive to price and so more success for the no-frills, low-price brands and Scarlet is very well positioned to participate in that growing customer segments.

And I think to date, we're quite satisfied to have this two-brand strategy where, one, you can attract the same price-sensitive customers; and with the other one, you are able to provide a full service, full content, full premium offer on the Proximus. Concerning your second questions on the postpaid, it's true that we have very good sales of postpaid 14,000 on CBU but also 16,000 on EBU.

I think you see mainly postpaid increase coming from our convergent offer, where we see good traction on the Tuttimus/Bizz All-In and that's of course including postpaid. We have also a product within postpaid that is today a bit -- a complementary project to the previous prepaid, it's called full control, so it's a full postpaid offer, but where you cut your prices.

And so in that sense, it helps people to control their cost spend. So we have the advantages of the postpaid prices that's also the prepaid capping in terms of subscription, so that's certainly also one element which is helping the drive of postpaid within CBU.

And then you have the normal Mobilus and Scarlet increase in postpaid. So various factor, but indeed, very satisfying results.

And perhaps, the last element I could add is that we have also changed a bit the way we do subsidized offer with the new type of subsidization which is independent from the subscription and that's also getting good traction from the customer as it gives the customer more freedom to choose for the phone they like and the subscription they like without any effect on the tearing for the time being.

Three questions. First of all, do you see that churn is picking up after the EC switch?

Secondly, when do you expect new cable wholesale rates to get implemented? And what is the kind of tariff that you would see as fair?

And then the third question is on the VOO. So we have -- extend of cable regulation, we have a new government in Wallonia.

Do you believe that the combination of these 2 elements could maybe result in a quicker sale of VOO?

So concerning the churn, I mean, I said it in my introduction that the churn has decreased in Q2. Of course, that's without EC switch, but I think we're quite happy that a lot of efforts that we've undertaken to try to improve customer experience gives -- improved the quality of our line, of the vectoring and of the set-top box, something like that, is currently really peaking into a reduced churn.

EC switch, it's still early days. We don't see major impact today of EC switch.

So I cannot really give you any indication on that. So far, no real impact.

The only thing which is annoying is that today on the cell sides, they are not compliant with the EC switch regulation and that's a difficult discussion we sometimes have with the customer where if someone comes from SFR, they expect us to do the whole procedures which we can't and we have to send them back to SFR and by the way, now Telenet. So that's the only annoying operational element you'll see on the EE switch.

Concerning cable regulation, to be honest, it's a bit difficult for us to give a lot of comments there. I think what you can read in the market analysis is that, indeed, the regulator wants to go further in access regulation for cable at what they call a reasonable price which is indeed going away from the retail minus, but it's not fully a cost-plus where [indiscernible] to be honest, it's difficult for us to tell because it depends very much on how which type of model they will use.

In terms of timing, it's probably more easy to say, I mean, the current consultation phase is going on until the 15th of September. After that, the BIPT will have to integrate some of the comments of the various players.

We expect them to finalize that by the end of the year. Then it needs to go to Europe, where Europe will have to look at the market analysis and will probably take at least 3 months if they accept it in the first wave but could take longer if they challenge the methodology of the BIPT.

And therefore, it's between bracket, I can say that we're very surprised by the way the BIPT is defining the market. As you probably have read, that they define the -- that cable market as being 1 market and the copper and fiber market being another market which is, of course, a strange way of defining markets.

We have seen in many other countries like in Spain, like in Portugal, like we also have indication in the Netherlands, that now regulator are looking at broadband markets as one market with different technology. In Belgium, to our knowledge, it's the only market where the distinction is made between a copper fiber market and a cable market.

So of course, easy to say that everyone has a significant market position when you tailor the market description to what helps you to regulate. So this will certainly be one element that we, at least, will put forward to BIPT, but I guess also Europe would put forward.

As today, in Europe, you'll see more and more of the commission trying to favor deregulation and more investments and I think the market analysis that is today on the table is not so much going into that direction. Concerning VOO, your last question, it's also very difficult to say because the new government has just been announced.

I think the major shareholder today in VOO is the [indiscernible] and the city of [indiscernible]. And to my knowledge, there is no change in the people that are running those 2 main shareholders.

So I think so far, I have not heard of any changes and any new movements in shareholding of VOO. But to be honest, we will certainly have more information on that in the months to come when the new Wallonia government will have settled.

I have a follow-up on that tax comment you made, very hopeful to get a sort of initial guesstimate for the ETR in 2018. I have a question though, what would it have been under the current, under the current overalls?

And the reason I'm asking is, from my understanding is that the reform also removes the notional interest deduction. Could you confirm that's correct?

I understand the notional interest sort of deduction is declining, but it's for the benefit. So I was just wondering how would 2018 look if everything sort of stay the way they are at the -- at this at the moment?

The second question I have is just regarding the FTE reduction that sort of incrementally came down a bit less in the second quarter than the first quarter. Could you comment on the phasing in the second half of this year for any further headcount reductions that you may be planning?

So on your first question, the ETR in 2017, as I said, without this, any impact of the reform would be close to 30%. And our forecast for the next year is that the benefit of deductions and the first one being notional interest deduction is decreasing because of very low interest rates.

So provided interest rates are, say, low, we were expecting 2018 to have an ETR which would have been north of 30%. And it's true that in the current reform, what's expected is that there are changes in the deduction, including a revision of notional interest reduction.

So all in all, that's our understanding and that's our current analysis. So we need -- as we get to 27.5% % to 28% is certainly a benefit compared to our expectation if there are no reform.

On your second question, so the full first half of this year, we benefited from two waves of departure of the early release plan which has helped decrease FTE by 832 FTE for Q2 and was roughly the same for Q1 because we had the benefit of the departure of the first wave which was in July 2016 and the second wave in January 2017. So as of Q3, we'll have in volumes just year-over-year, 1 wave which is roughly half in terms of number but we will also have in our wage indexation that starts in July 2017.

So you when you will be comparing seasonality for the balance of the year, there will be a lower volume impact, but there will also be a wage increase which will then diminish the impact of our workforce cost reduction for Q3, Q4.

Just a follow-up. Do you have plans that you can talk about already to reduce the headcount further in the second half of this year?

Well, the key drivers are beyond the -- early this plan, natural attrition and in pension. So beyond that, that's how we operate.

Just three quick questions, please. On the new More4More offers which will apply from the 1st of August, presumably, your existing customers have all been notified by now of the changes.

I just wondered, have you seen any pickup in churn over the course of the last few weeks? I mean, we're obviously seeing price increases on fixed and mobile landing less well not just in Belgium but across Europe.

And then second question, you mentioned fierce competition in enterprise. I was just wondering which of your competitors are being the most aggressive in that segment?

And -- or Proximus losing share in the Enterprise segment as a result? And just lastly, if the BIPT market analysis goes towards regulating your fiber network and current investment model which you don't think is acceptable, could that actually lead to you scaling back your fiber plans that were announced in December last year?

So first on your first question, on the -- if -- do we see some churn after our More4More announcement? I think in the first two weeks, yes which is a normal pattern when you announce a price increase, you have some customers that are a bit more price-sensitive, they tend to leave.

So we see a bit more port out in the first week, but that has already normalized today. So I think it's a normal pattern from a price increase.

And I think it's important longer run that we find a way to monetize the investment that we do. We've done quite a lot of investments to support the data usage.

We want to offer our customer the possibility to use more data in Belgium and abroad. And I think a lot of customers are really appreciating that.

But we also need to be realistic and to drive some price compensation for that and that's what we have done. And we will see how the market will react, but I think today, this is still the way we try to manage the business by improving our network, improving of services and trying to monetize that in the best way and mainly with the More4More strategy towards our customers.

And we also then of course, have the alternative on Scarlet for the people that are not willing to go to the premium brand. And I think that model and that equilibrium so far has worked quite well for Proximus and I don't see why it would fundamentally change the paradigm.

As far as your second question, so it's true that we have -- that we're seeing increasing competition in the enterprise market. It's not a secret that our competitors are targeting the enterprise markets because we're pretty strong in that area.

Who are the competitors? Well, it depends on, I mean, if you talk telco, then you have the traditional ones which is -- which are cable being Telenet and VOO; and in mobile, you have, of course, also Orange.

Next to that, you have, of course, over-the-top players that start to impact also the enterprise market. If -- regarding your question on, do we -- are we losing share?

For the time being, what I can say, we're not losing share. In the contrary, like in mobile, we're still gaining market share; and for the rest, we're sustaining as much as possible.

Then on your last point concerning the BIPT, but to be clear, I mean, today, on the potential fiber regulation, the modalities are not yet defined. So the BIPT has put forward the full potential modalities, of which one is completely in line with the current deployment model of Proximus.

Three are slightly different. So what we had said is that, indeed, if we would be imposed a completely new topology for the development of fiber, that would not -- that would increase cost of deployment substantially, that would also delay the deployment.

But first of all, we would try to fight that because our objective is to offer Belgian customer the strong alternative with a gigabit network as fast as possible and for as low cost as possible and both cost to build and cost to run, so that's really our objective. If we would end up in a model, but to be honest, I would not understand very much why it would be the case because I think there, the BIPT and ourselves, we have the same objective, is to offer in Belgium an alternative to a gigabit network as fast as possible and at the lowest cost as possible and we really think that our topology is the one that deliver that.

If in the unlikely event it would not be that and we would have to invest substantially more and delay investment that would indeed force us to look back at our plan at potentially scale back our Fiber-To-The-Home investment plan. But today, we're not saying that that's the case.

Today, we're still investing and we hope that we will be heard for the model we have put forward which we really think is the best development model for Belgium, the most efficient and the fastest to get a second gigabit network available for the Belgian customer.

Referring to your second question, I realized that I forgot to answer the ICT part, so I gave you the competition for the telco part. In ICT, of course, there the competition is much more scattered, depending on the activity.

Meaning, if you talk security, competition is different. Then if you talk cloud or if you talk UC, so there we have a much more scattered landscape.

Two questions, if I may. The first one is I can see that you point to the -- achieved for the consumer market from [indiscernible].

So could we read anything there? Any -- is there any change in your ambitions in terms of a reasonable approach to content?

And my second question is regarding TeleSign. Is there anything more that you can tell us there, perhaps, in terms in terms of synergies?

And just a very naive question there, why actually is TeleSign going to be included within BICS? I can't understand the network answer.

But wouldn't that make sense actually to integrate to the [indiscernible] with the business teams of Proximus?

So on your first question, I think Giom Bhutta [ph], will indeed join the company by the end of August to take care of the Consumer business. Giom has a lot of experience in telco because he has been working quite a long time at SFR.

He's recently at Canal. But so for us, I think it's a very strong addition to our Executive Committee with someone that has broad experience, both in the telecom and the content.

Would that change our approach to content? To be honest, so far, not.

But Giom is not yet there. So I mean, he first needs to come and to see.

But so far, on content we really try to remain what we call, an aggregator of content, offering our customer as much choice as possible, so very important local content, but also more international content, sport content, kids content at the lowest price. And I think, today, that remains our main objective, building, of course, next to that also a more personalized and targeted content which is something that we're still developing and where indeed it could be that Giom can really help us growing faster there.

This is Daniel. On TeleSign first, we're a days or a few weeks away from closing.

We've not closed the transaction yet. So we're ready to activate synergies and hopefully, to give you more color next quarter.

Now why did it make sense to be part of BICS? Well, TeleSign generates billions of messages to users across the world.

And by integrating TeleSign and we've been explaining we're creating what we call the first -- the world's first end-to-end communication platform as a service. So from the digital customers of TeleSign using APIs and the authentication services to all the mobile network that we reach directly with the big network.

So it is a very complementary segment to our other business segments being the traditional wholesale segments and that's why it's part -- well, BICS is part of our transformation, addressing the communication needs of the digital market.

And first, just to add to that, I think it's very important that it's integrating BICS, but it does not prevent us to try to find synergies and potentialities to develop the technology of TeleSign within the Proximus environment. There are already some contacts with our affiliate in Luxembourg which is quite strong in fin tech.

So I think the integration will be in BICS because it's quite obvious, it's an international play. But of course, we will also try to leverage the knowledge and the technology and the competence of TeleSign for our Proximus and other businesses.

My first question is on the direct margin. You also -- already mentioned that this mix and the roaming impacts.

But could you talk a bit more about it? Because I think that the mix and roaming has always been the case and if it's quite a certain step down that you see in that direct margin?

And the second question is on the cost cutting. It seems that you are already running well ahead in the first half of the year.

Also last year was pretty well ahead. What is the idea there?

On that €50 million of cost cutting, was it more front-end loaded? Or are you simply running ahead and there might be opportunity to do a bit more?

So concerning the direct margin, I think roaming has some impact on direct margin. I think we will see probably even more impact next quarter because the impact of the new Roam-Like-At-Home in Q3 will be more than what we have in Q1, Q2.

But indeed, the direct margin lowering is a combination of mix. It's, I mean, less fixed voice.

I mean, that's I think the main impact on our direct margin. A bit less of ICT projects this quarter and also our wholesale which is also an impact on the direct margin.

So we -- I don't think we can relate the direct margin directly to roaming. I think it's a combination of various elements which are typical to telco today which is a transformation of historically, a lot of fixed voice to more data, to more TV, to more ICT and that of course, there's an impact on the margin.

And that's why of course, as your second point, the cost cutting is very important because we need to find ways to run our business with less OpEx. But I think there, I will let Sandrine give you a bit more indication on that one.

So it's true we're tracking well on our cost decrease ambition. We're even a bit more front-loaded as you said.

Now we've not revised our ambition for the years of decreasing our total FX by a net €150 million which we still are working on all various initiatives, the full and the whole dimensions of the company. And we will see when we revise our next 3-year plan.

But so far as you know, this €150 million is the result of a strong cost decrease by €400 million of cost initiative. At the same time, we also have some increase which are either indexations, either volume or either driven as well by new activity.

And one of them being the fiber rollout which also comes with not just CapEx, but also OpEx. So I think it's important to have in mind this balance between the effort we're going on the gross net saving and the activity of the company which is still growing fast in some elements on the volume point of view and some indexation.

That's clear. And then maybe as a follow-up to translate that into the outlook for full year '17, you're guiding for the -- you had nearly stable or Domestic revenues and group underlying slightly growth, you're running well ahead on that.

And I think then the cost savings will be even be additional cost savings in the second half of the year. And roaming will have some impact.

But not that much extra versus the first half of the year, if you're looking on your slide on Slide 18. What are your thoughts about that?

Is there's something that's going to be a more difficult comparable base for the second half of the year?

So for the second part of the year, as I said on the OpEx side, we would not get the benefit of what we had for the first half. With 2 waves of E&P, we'll have the wage indexation, but as well, roaming in Q3, we expect it to have a bigger impact that's what we've observed on the first half of the year on our direct margin.

And I think you should look at the sequence of our Qs in our quarter in 2016. We had an excellent Q3 EBITDA in 2016, so that would be a tough comparison.

Just in terms of expectation for the balance of the year Q3, Q4, we really had a very high Q3 2016 where we massively decreased our costs. So in terms of expectation for the next quarters, that's something that you have to take in mind.

And so that's why we reconfirm our guidance for the full year.

Okay. Since there are no further questions, I think we can close the call now.

So thank you very much for calling in. Should there be any follow-up questions, you can contact the Investor Relations team.

Thank you.