Good morning, ladies and gentlemen, and thank you for standing by. Welcome to Liberty Global's Third Quarter 2019 Results Investor Call. This call and the associated webcast are the property of Liberty Global, and any redistribution, retransmission or rebroadcast of this call or webcast in any form without the expressed written consent of Liberty Global is strictly prohibited. At this time, all participants are in listen-only mode. Today's formal presentation materials can be found under the Investor Relations section of Liberty Global's website at libertyglobal.com. After today's formal presentation, instructions will be given for a question-and-answer session. Page 2 of the slides details the Company's safe harbor statement regarding forward-looking statements. Today's presentation may include forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995, including the Company's expectations with respect to its outlook and future growth prospects and other information and statements that are not historical fact. These forward-looking statements involve certain risks that could cause actual results to differ materially from those expressed or implied by these statements. These risks include those detailed in Liberty Global's filings with the Securities and Exchange Commission, including its most recently filed forms 10-Q and 10-K, as amended. Liberty Global disclaims any obligation to update any of these forward-looking statements to reflect any change in its expectations or in the conditions on which any such statement is based. I would now like to turn the call over to Mr. Mike Fries.

Great. Thanks, operator, and welcome, everybody. I appreciate you joining the call today. We've got plenty of ground to cover, as usual. So we're going have to try to keep our prepared remarks a bit shorter today so we can spend more time on Q&A. And as usual, I've got a number of execs on the call with me, who will be sure to get engaged as needed in that conversation. So I'll kick it off with some Q3 highlights. By now, you've combed through our financial and operating results, no doubt, and Charlie is going to dig into those in a moment. From my point of view, there were 3 key takeaways here. First, we're tracking to guidance, in particular, free cash flow. And as we've discussed all year, that's a function in part of resetting our capital intensity levels. With P&E additions down considerably, we've reported 40% rebased OFCF growth if you include Switzerland and 80% if you don't year-to-date. Again, that's a major step-up from prior years, and it's happening across all of our key opcos. And then third, we are absolutely descaling central costs, both at corporate and in our centralized T&I operations. Several of you have asked for clarity on this point, and Charlie has an entire slide on it coming up. The punchline, though, is that the costs we incur to fulfill the service agreements we've signed with Vodafone, Deutsche Telekom and our Dutch JV will more or less equal the revenue we're generating from those agreements as they rise and fall. So there will be no significant mismatch in those costs, which means the additional costs to support our remaining operations, which is largely allocated to our opcos, will not be impacted by these TSA agreements and should, in fact, be flat to…

Thanks, Mike. For those following, I'm now on the slide titled Group Overview. Revenue in the quarter was broadly stable excluding Switzerland at minus 0.2% and with Switzerland at minus 0.6%. Telenet continues to be impacted by the loss of the Medialaan MVNO contract. And we also started to recognize TSA revenues from the assets we sold to Vodafone in the quarter, which contributed $26 million of incremental revenue, but we don't include that in our rebased growth rate. OCF declined 4.1% including Switzerland and 2.9% without it. This was in line with our expectations for the quarter. The U.K. was impacted by the year-on-year increase in network taxes, which will continue to increase but at a lower rate over the coming years. We also saw an increase in programming costs related to prior year renewals, such as BT Sport and UKTV and the recent multiyear Sky agreement. Swiss OCF was impacted by cost saving and continued investment in key revenue and growth initiatives such as the EOS box rollout and the 1-gig upgrade of the broadband network. OFCF continues to grow strongly and was up 63% excluding Switzerland and 35% with it. Remember, the Swiss numbers are impacted by the investments in those initiatives such as EOS set-top boxes and the 1-gig upgrade, which we would expect to normalize over time, resulting in a lower CapEx figure. Liquidity remains very strong with cash sitting at $7.4 billion. Now when we completed the sale to Vodafone in July, we reported net cash proceeds of $11.3 billion, which, together with our Q2 cash balance of $1.3 billion, came to $12.6 billion. The $5.2 billion difference is made up of the $1.6 billion UPC term loan repayment, the $2.7 billion Dutch auction tender completed in September, as well as a combination of…

And with that, over to you, operator, for questions.

[Operator instructions] We'll go first to Michael Bishop with Goldman Sachs.

Firstly, on the U.K. RGU growth and the wider performance. Mike, I think you said it was the best quarter in four quarters for Lightning net adds. So sort of backing that out, given the relatively weak overall RGU trends, particularly TV, I was just wondering if you could give us some comments on what was happening in the existing footprint outside of Lightning. And then secondly, on the free cash flow guidance, it looks like if you broadly take the $700 million to $750 million, removed Switzerland and then sort of removed the higher dividend from VodafoneZiggo and the Telenet free cash flow, you're basically at 0 free cash flow. So as we look ahead to 2020, I was just wondering if you could walk through the moving parts to give us a bit more color on how effectively the ramp, including Virgin Media, could improve cash flow generation and where you think the group might be able to head to in 2020.

Sure. Thanks, Michael. Charlie, you'll, I'm sure, prepare an answer to the free cash flow question, make sure we understand that correctly -- we understand that one correctly. And I'll ask Lutz to chime in, of course, on the U.K. But in principle, as you correctly said, and as I pointed out, the RGU growth in Lightning is pretty steady. If you look at total net adds, broadband net adds, telephony or video, we've been pretty consistent in the last four or five quarters, and this was a better quarter than we've had in the last few, marginally better. But I think it demonstrates that the product matters, the Lightning project is delivering the sort of penetration rates, ARPUs and returns that we expect. And we do understand that people like the detail of Lightning. We were just trying to make the slides a little shorter today, but we weren't avoiding any disclosure on that. Lighting continues to be a terrific program for us from almost every perspective. In terms of the -- in the BAU or GAU, depending on how you define it, and I think Lutz will dig into this and I made comments in my remarks, it's a competitive market. That's for sure. Particularly, we are focused on more of the premium product and the higher-end product, especially in video. So we're trying to retain customers that we view as profitable customers and the kind of Virgin customer that means a lot to us, which means you're going to lose some customers at the lower end. And that clearly happened in the BAU market, in particular in video. It was not a great quarter for us in video. I think -- I'm not sure we're disclosing the details of that, but you can probably do the math.…

Just on the free cash flow, and I appreciate your numbers. If I just break it out into the key groups. I think your point is, essentially, if you take out Switzerland and use the $600 million number and then you subtract Telenet, you'd probably end up with only around $200 million and then you're saying that the $200 million effectively equates to the dividend. I think what I would point out is when we talk about shareholder distributions from Holland, which are about €600 million, remember that €200 million of that, or €100 million are our share. So €300 million is our share. €100 million of that comes in the form of a loan repayment that isn't a hit into free cash flow. Next year, we're not probably making a loan repayment so that kind of repayment will be cash available to fund the dividend. So that's one point. I think the second point is a point we've tried to emphasize on a number of occasions. The UK on a net basis isn't generating a lot of free cash flow. But there are 2 very distinct businesses. You have the Lightning business, which, depending on how you look at the allocation of debt, is a negative $450 million free cash flow business, whereas you have a core business that is generating something like $600 million, $700 million, $800 million of free cash flow. And that is why there's a distinction. So if we stand back and think about our assets, we've got a highly generative asset in Belgium, very generative asset with upside in the Netherlands. Switzerland actually is a depressed level at 24%. Traditionally, it's been a lot higher, but we're high in the investment cycle. And then you got this Virgin business that splits between a very generative business as usual, but a highly investing Lightning business. And then the delta is clearly Eastern Europe and corporate. Eastern Europe is slightly positive on an aggregate basis, and the corporate net number is this $200 million that we've been talking about, which we tried to explain in the slides. I hope that's helpful. If you have any other questions, maybe we could follow up off-line with Matt and the guys.

We'll move next to Jeff Wlodarczak with Pivotal Research Group.

It sounds like the price hike accounted for most of the UK RGU result. Do you feel comfortable you'll be able to return to maybe modest RGU growth in the fourth quarter? Or is that too high of a hurdle? And then how interesting is it, there's a lot of competition at the low end. How interesting is it to potentially launch a flanker brand targeted at lower-spending users to boost RGU growth?

Lutz, do you want to get this?

Yes. So right, I said that 25% is due to our conscious decision on video strategy. You can expect that somehow to continue. And then I wouldn't say that the market will repair itself from one quarter to the other, right? So competitiveness is still high. I mean we are not making a guidance for next quarter. But I think when you look at the numbers to, you come, I'm sure you'll come to the right conclusion. Your question on second brand, well, let's put it that way, right? So the more we invest in our network, the more also we want, obviously, to leverage that network. We have with Virgin Media a high-value brand. So I think everybody else in the market has also low-value brands, Plusnet, Now broadband. So you can be assured that we are working on that, yes, but we want to have this surprise on our side.

The one thing, Jeff, you will see in the fourth quarter is the ARPU itself should be better. So ARPU and revenue should be better because we'll have a full quarter of that price increase.

And we'll move next to Vijay Jayant with Evercore.

It's James Ratcliffe for Vijay. I just wanted a couple of questions, one on the operations side and one on the strategy side. Can you talk about what your thinking is around the Dutch JV? I know some of the put call options are starting to come up. And whether it makes sense for that to be stand-alone or potentially become a closer or more distant part of Liberty Global? And secondly, I just wanted to understand on the corporate costs and make sure I'm getting this right, that it looks like you're saying that once the existing TSAs fully roll off, you're looking at a net cost of around $500 million a year for corporate plus T&I as allocated to the currently ongoing businesses.

Charlie, you work up an answer to the second one. I don't know if those numbers are right. But with respect to the Dutch JV, James, we're happy where we are in the current situation. Both parties do have rights with respect to listings and then down the road, other sorts of liquidity provisions. But I would say -- and I'm pretty sure Vodafone said the same thing. We're very pleased with the performance and the progress that the joint venture is making. As I mentioned in my remarks, it's returned to growth on the top line and the OCF line. It's got great OFCF and free cash flow margins and a business that's generating dividends to us. So we're pleased with the progress of the venture. We had the whole Board there just a month or so ago. And I think Jeroen and the team has done a terrific job, and they're proving out yet again, in yet another market, the benefits of fixed-mobile convergence. Synergies are on track. There are some outstanding strategic matters around spectrum and wholesale access, but the fundamentals of the business look terrific. At this point, I would assume we are going to own our 50% stake. There doesn't appear to be any huge momentum either direction to drive a change in that. Charlie, do you want to address the corporate costs?

Yes. So just on Page 8, James, I mean, you have it right. So just to summarize, the T&I costs, we do cost on behalf of the companies we've either sold or ventured and then for the retained operations. And as you correctly point out, the retained operations were $69 million of T&I costs. $70 million in Q2 and $75 million in Q3. We're saying to you that we think that number should be flat to down over the next 4 or 5 years. And the reason is because we're able to flex the cost as the other TSAs roll off because a large amount of that spend is actually not internal labor. It's third-party contractors, and it's a sort of a variable cost model. And then you're also correct that the net spend, which is the classic corporate overhead, which is HR, finance, strategy, development, et cetera, etcetera, that was $55 million in Q1, $53 million in Q2, and it's $44 million in Q3. And we're telling you, that's going to be around $200 million in 2020. It'll probably stay flattish thereafter. But certainly, we wouldn't expect that to increase, and maybe there's opportunities for savings.

We'll hear next from Nick Lyall.

This is Nick from SocGen. Just two, please, on the U.K., please, Mike. It's -- you mentioned again the U.K. plans on fiber maybe for the next 7 million to 10 million. So it looks like BT is probably going to have to wait about 12 to 18 months before it launches its own full plans, if at all. So how quickly could you launch for these -- for the 7 million to 10 million? Is this something where you'd prefer to wait to see what BT does, what the regulator says? Or is it something that you think is maybe a bit more imminent? And then secondly, that means that U.K. -- it's already quite a complicated business with Lightning and non-Lightning anyway. It becomes more complicated. So do you think you'd have to think of different structures, different disclosures so that the market could value it properly?

Good questions. In terms of what BT will or won't do, I think when Phil speaks to the market, he's quite determined that it -- that he needs to invest in fiber. I think his challenge is the amount of time and capital it's going to take doesn't sync up particularly well with his financial position today, whether it's EBITDA, free cash flow or dividends or pension plans. So I think he's got a bit of a challenge in front of him to accelerate and rapidly roll out fiber. So I think you're right. It's going to take time for them to get any kind of momentum. But we do assume, for what it's worth, we do assume that BT will build fiber over some period of time. If you're looking at a 5- to 10-year time frame, we do assume that they'll reach some measure of critical mass. Certainly not the entire market, well south of that. But they'll pick up speed, and they will build fiber. They'll need some regulatory relief to do it, and they'll need some financial relief on some basis to do it. But we assume, just to be clear, that they're -- that BT will build some fiber in some portions of the markets. As you point out, there are 7 million to 10 million homes that are -- we could possibly go out and build more quickly and potentially more efficiently. And while we're pleased with Lightning at the 400,000 to 500,000 home a year clip, we are unlikely to do that through Lightning on balance sheet and impacting our free cash flow and taking a fair amount of capital to do so. So I don't anticipate that you'll see Lightning on steroids, where we take the -- at least not in its…

We'll move now to Ulrich Rathe with Jefferies.

One short. I have one question and maybe just one short, very short clarification. The question is, what is the hurdle? What are the decisions that need to be made or what's the evidence you're waiting for to move forward on the shareholder return decisions at this stage? You're sort of repeating for quite some time now that it's all a question of sort of generating value, and I think that's fair. But in terms of sort of trying to understand what the time line is and what it depends on, it would be interesting if you had anything in mind that you can point to, what needs to happen or what you're waiting for or what you still have to decide. And then a quick question for Charlie for clarification. In Central and Corporate, there was a bit of a revenue pop. You highlighted the costs were more or less flat quarter-on-quarter. Was there a revenue pop? In the report, it said there were high CPE sales to the Netherlands. Is that all that is, this step-up? And then that looks a bit like a one-off. Could you just confirm?

Thanks, Ulrich. On the buyback, I mean, I believe, and Rick will check this, with the $3.2 billion in shares we've repurchased this year, this is the largest amount of stock we've ever purchased in a 12-month period and not an insignificant amount of stock. So I think anybody who's looking at their own shares and trying to manage a buyback program wants to be measured and thoughtful about that. We did just complete a $2.7 billion Dutch auction tender at $27. We were fine purchasing at that price. And today, we're at $24, $25. So you want to be thoughtful about the timing of these sorts of things. That's point one. Generally, if you go into shares a while, and you have, and follow them for a while, we provide some measure of guidance around buyback activity on our year-end call. And that's coming up, of course, as we finish the year. And so I think there's no particular science behind it. And then thirdly, I suppose the last thing would be we're always evaluating opportunities to create value for the Company on both sides of the spectrum, buybacks and investment. And we're looking at those in relation to one another with a fair amount of discipline. So that as we evaluate opportunities to put money to work in an industrial way, we want to be sure that we're able to demonstrate returns that would exceed what we could do just by buying our shares. So I don't, I wouldn't take our lack of a buyback program at this point in time or the fact that we haven't announced another Dutch auction as an indication of anything, really. Just that we're pacing ourselves as we should. And as we take measure of the year, we've put quite a bit back into the stock. $3.2 billion is not a small number. And I think we're taking a longer view of this, and perhaps maybe a bit more patient than other shareholders.

Just on the corporate side. You're quite right. Yes. So we do buy set-top boxes on behalf of VodafoneZiggo. The reason why we very specifically on our central update focus you on OpEx and CapEx is that is not in those numbers. So it's a cost of goods sold line is the way. And then the second thing to say is that is essentially a zero margin -- slight positive, zero margin-type activity. And it's a hard number to predict because it's quite volatile because when you do set-top boxes, there's 2 things. One is you can refurbish existing set-top boxes and one is you can buy new. The reason they're buying a lot of new set-top boxes is they're rolling out the new EOS platform, which, by the way, has been a terrific success. Generally, we see a lot of our boxes being refurbished because they have good average life and they work pretty well. In fact, that's one of the factors in the slight decline in OCF in some of the markets because we've shifted less from new box sales to refurbished box sales. Refurbished counts as OpEx under the magic of U.S. GAAP, whereas new box sales are CapEx, which is one of the reasons why OFCF is probably a more accurate metric of what the underlying cash flow is doing than the old OCF. In terms of predicting how it will go, I can't really predict. I mean if they continue to sell like hotcakes, we might see an acceleration in EOS, we might not. But just to -- it may slightly distort the revenue. But on a cash flow basis, it's basically a breakeven number.

Our next question comes from Carl Murdock-Smith with Berenberg.

Just one question from me. I mean I track very carefully the churn, and I don't think you've reported U.K. churn this quarter. So I was just wondering if it was possible for you to provide an update. I think the 12-month trailing was 15.0% last quarter.

Yes. I mean, historically, we realized that none of our peers report that churn figure either, and so I don't believe that the lack of reporting -- and Charlie, you'll have to jump in here, and Lutz. I don't believe we're not reporting it because it's not a good number. I think we probably made a determination that the amount of detail we're reporting in our markets wasn't consistent and not consistent across the industry. But Charlie or Lutz, do you want to address that? Lutz Schüler: Yes. I said, right, when we -- when I was referring to the RGUs that 75,000 are due to the pull-forward of the price rise, right? So we are predominantly selling triple. So I think you cannot make up the number ourselves. On the other hand side, right, I mean, I try to get the intelligence from the competitors and I get less and less. And why would we disclose more and more?

Okay. I suppose I'll just add on that, that BT does provide fixed churn every quarter. And I suppose just as a follow-up, it's interesting that you've lost TV RGUs. But at the same time, Sky this quarter has also turned to declining RGUs this quarter as well. So where do you think those TV customers are going? And do you think it's evidence of more cord cutting in the U.K. TV market? Lutz Schüler: I think or we know that these TV customers are the low-end customers. What do I mean with that? These are predominantly customers who have -- used to buy our triple-play entry product, right? So paying roughly, with time-limited discounts, £30. And then after 12 months, when they get off the time-limited discount, it's a very high jump. And then they simply value the pure video product not as much then to pay more for it. And then you see high churn. And therefore, I can only speak for us. So therefore, you see these 25,000 on a quarterly basis are pretty much that. High level -- the high-end packages actually are very stable. So -- and they provide also obviously higher margin. So therefore, I would not really see the world turning very much to an app world, a pick-and-choose world. So I don't see that accelerating. But obviously, we have to watch that carefully. Over a longer time period, we will see that trend. But these numbers are not really referring to that.

Robert Grindle with Deutsche Bank has our next question.

Yes. My question is about the U.K. deal you've just done with Vodafone. Congratulations on that. You had another two years to run, though, on your existing MVNO with EE, so it seems very premeditative to strike a new deal so soon. So I wondered, what drove the time line on that? And what was the swing factor? Was it 5G access having Vodafone as a customer for Virgin business? Or was it the financial terms or something else?

Well, I'll let Lutz dig into it. It was -- from my point of view, it was all of the above. There was no other strategic purpose to it. But to switch providers is not a simple thing, even though we have a core network that allows us to do it relatively seamlessly. From a consumer's point of view, you need to get prepared for these. So I think it was smart for the team to determine well in advance which direction we're going, to put that out to bid, if you will, and as a result, derive a fantastic contract that, as we pointed out, guarantees 5G access, which we were not getting in the other deal, and has much better economics. But I wouldn't read anything strategic into it. That would be my observation. But go ahead, Lutz, if you want to add anything more. Lutz Schüler: Yes. Exactly. We've done an RFP in the market, and we wanted the best technology at the lowest price, and this is what we found. And we have the opportunity now to launch 5G as fast as we are able to implement it, yes? So we are not sticked to the end of the contract in the 5G world.

We'll move now to David Wright with Bank of America.

A couple from me first. The first one follows on from Robert and the answer you've given. My understanding is that 5G services, you can migrate instantly away from the BT MVNO onto the Vodafone contract. I think you have like a clause that allows you to exit. So is it possible we could see customers coming before that year-end 2021 expiry? And then beyond that, I think you mentioned, Mike, a little earlier on that we could infer what we wanted on perhaps this Swiss deal. It's obviously something that I think the Sunrise shareholders maybe saw strategic value in, but didn't necessarily like the means to raise capital or the price paid. Is this possibly something where you could even sort of migrate to some kind of earnings earn-out clause so you potentially take less capital upfront? You're not exactly desperate for cash right now, of course. If you don't mind me just following up, I wonder whether -- there's a reciprocal agreement on some network usage, backhaul, et cetera. Is this something you guys could ever even extend to reciprocal wholesale access to Vodafone customers or even any wider wholesale access to the VMED network?

Well, the -- I mean, we have -- go ahead and address the first point, but not the second one. Go ahead. Lutz Schüler: Yes, exactly. Obviously, right, we have now really engaged in a closer partnership with Vodafone. And as we have said, right, Vodafone very fastly deploy 5G base stations, and they need a backhaul opportunity. So therefore, we have also closed a huge deal on dark fiber and mobile backhaul at the same time when we have closed an MVNO deal, and we are very pleased with that. And in general, so dark fiber, IRU, mobile backhaul deals, this is an area which has huge growth opportunity in the future. And therefore, we focus more and more on this kind of business.

I think with respect to the second part of your question, which is wholesale access of the end-to-end network from a fixed retail point of view, that is not something we've discussed with Vodafone. That's obviously a very different conversation and I think one that we'd have to take a longer view of. And at this point, nothing to add. I mean we don't think that's necessary in this market. It's certainly not legally or regulatorily required, and I don't see us jumping into discussions of that without broader considerations.

We'll hear next from James Ratzer with New Street Research.

I had a question and a very quick clarification. Just regarding your -- in the U.K., customer growth on your BAU footprint. I was wondering if you could kind of talk us through what your strategy is to try to return that customer base to growth and actually, indeed, your strategy to return it to growth. Or are you going to be much more focused on trying to extract revenue growth there just through price rises alone? And then just a clarification on the churn point. Lutz, I think you said 75,000 from that. I want to check my math. That's around a 1.4 percentage point impact from churn in the quarter, which will be about kind of 5 points annualized, implying annualized rate of churn to step up to about 20%. Is that a fair assessment based just on the Q3 churn levels?

Lutz, do you want to address those two questions and I'll chime in also? Lutz Schüler: Yes. So I think on the churn number, right, I told you the RGU number. So you have to turn it into customers. But on an annualized churn, you get a bit like 0.5% on it. Then you're right, right? And in the quarterly churn, obviously, the increase was higher.

But it was the quarter of the price increase. Lutz Schüler: Yes, it was the quarter of the price increase.

It's always moving higher. Lutz Schüler: Yes. And we are getting back to it. Exactly.

And I think if you -- if we would disclose it, as you calculated it, it would be -- would not be an aberration. Go ahead, Lutz. Lutz Schüler: What was the first question again? Sorry for that.

Yes. So first question is about what the strategy is [indiscernible] customer increase in BAU or whether it's more going to be focused about price on an ongoing basis. Lutz Schüler: Yes. Okay. So right, this is where our consumer strategy kicks in. So we started that a year ago, and we have four growth levers. One is fixed-mobile convergence, right? So we started that. We know that fixed-mobile converged customers have a substantially lower churn, 2, 3 percentage points. So therefore, we drive that. And when you see our mobile net add numbers or the accelerated number of fixed-mobile converged customers, you see that we will get an improvement from that over time, yes? So we have started at 19.6% a year ago, and now we're at 20.7% or something in that area. And that is the second quarter after we launched. So you can be assured we will push for much more, yes? And then take the MVNO deal with Vodafone into account. So that is a very strong weapon. That's number one. Number two is the SOHO business, right? So we have put now the small office/home office business into the consumer machine. So we can leverage all sales channels, all marketing online, and it's huge -- it's much less penetrated areas. So our market share in SOHO is substantially lower. So therefore, we ramp up that. Number three is do more regional sales, yes, so we are getting more data analytics in place. MDU as a segment we want to target more, and we are doing more and more on that. Get to higher market share in several regions. Greater London, our market share is 36%. Scotland, our market share is 60%. So we are focusing on that. So therefore, I'm not -- we are not giving a forward-looking guidance, but you can be assured that we have put a strategy in place to get substantially better in the GAU business, but these things take time.

And you think that can be done without price discounting? Lutz Schüler: I mean price discounting, I mean, that is in general, right? I think discounting has increased in that market, yes? So I think that's a factor of market saturation, right? So -- and also, more customers are within the minimum contract length. So that means everybody is discounting more to get a higher share on gross adds. And so therefore, simply, we need to offer value for money, right? And that means we are in a position that we need to offer discounts as well if we want to get a certain number of gross adds. I mean, obviously, we will see a game changer, an end-of-contract notification coming from February onwards, and let's see how the market is evolving there then in terms of discounting. So if the market is then -- right. Because what is happening is you give a high discount within the minimum contract length. After that, the customer is then asked to pay more. What the customer does is calls in and asks for a retention discount to stay on that discount, right? So with an end-of-contract notification, obviously, we talk about that even more. So I think it's more customer-friendly and more market-friendly when we are getting a bit more rational here.

We'll take our final question from Maurice Patrick with Barclays.

Maurice here. So quick question on Switzerland. I can see how you're clearly very disappointed about the deal failing. But you do sort of highlight in the release that you do see, I guess, early signs of a recovery there when you point to improving NPS trends and subscriber trends. I mean OCF remains pretty negative. I wonder when we'll start seeing the impacts on the financials and the improving operational trends that clearly you can see already coming.

Sure. And Severina is online here, she's welcome to chime in. I mean we're tracking a whole series of operational metrics there. RGU movement, of course; ARPU, which is growing every quarter; the rollout of EOS boxes so that we can get our base to about 50% advanced set tops by the end of this year; 1-gig homes and 1-gig penetration and seeing the impact of the faster speeds in the marketplace. Now we're investing quite a bit of money to achieve that. So on one hand, you've got some really positive metrics on customer, on the customer front in terms of NPS and net add performance and ARPU. On the other hand, we're putting capital into products, into technology and also into digitizing the operations themselves. So this year, I think we were clear that this is a year of investment. But we want, but we would want to see metrics improve that I've just described, and they have, and they are. And that through the course of next year, we'll start to see the pendulum swing the other direction. We're not providing guidance as we sit here. But I think we've always said that 2020 is an important year, an important inflection point for the business if you keep hitting on all cylinders. And so far, we have. Severina, would you like to add anything to that?

No. Yes, Mike, you basically pointed in the right direction. As you said, this year was a year of investment, and we actually worked on multiple areas. Rolling out EOS, doing the 1-gig, investing in Simply Digital. And as the organization also mature, we expect to do all these activities faster and better as well as being that the efficiencies coming, paying off from the investment we do on Simply Digital. Next year, we should see, yes, the organization getting more mature and benefits coming through to a greater extent than this year.

Maybe to wrap it quickly, I think there's 3 points. One, you can tell from the conversation today and from our remarks that we are focused, first and foremost, on pushing our operations forward with 2 primary goals: one is profitable subscriber growth; and secondly, free cash flow margins and yield. And all of these operations, even Switzerland, still very profitable on the free cash flow line. And our view is that each of the businesses, when you aggregate them together, are free cash flow machines, really. And that's changing the narrative a bit for you, and something we will continue to do a better job of is getting you to rethink the narrative of our operations. And I think you should expect more of that from us. Strategically, we're always looking at ways of recognizing and crystallizing value, both in our operations and in maybe new operations. So you should expect that we are very busy and looking at ways to both crystallize and realize value in our existing markets and possibly others, but always doing that with an eye towards returns and value creation, nothing else. And then lastly, we continue to look at our capital structure. And of course, I think our credibility on buybacks and the commitment to buying stock is a question. So it's just -- I can appreciate the need or the request for more disclosure on that, but just be patient. We're always looking at those things in parallel. And so we'll have probably have much more to say about that through the course of this year and the early part of next year, and we appreciate your support. Thanks, everybody.

Ladies and gentlemen, this concludes Liberty Global's Third Quarter 2019 Results Investor Call. As a reminder, a replay of the call will be available in the Investor Relations section of Liberty Global's website. There, you can also find a copy of today's presentation materials.