Millicom International Cellular SA (TIGO) 2012 Q2 法說會逐字稿

Hello. Welcome, everyone, to this Millicom half year results presentation. My name is Justine Dimovic; I'm Head of Investor Relations. And I hope you all found the slides that we made available today on our website, www.millicom.com.

Before we start, I would like to remind everybody that the safe harbor statement applies to this presentation and the subsequent Q&A session.

So with me today on stage, I have our President and CEO, Mr. Mikael Grahne; and our CFO, Mr. Francois-Xavier Roger.

Without further delay, I will now hand over to Mikael, who will present our year to date highlights, and take you through our strategic update.

Good afternoon, everyone. It's great to see some of you here today with us in London. Let me start with the highlights of the first year, if you turn to slide 5.

In H1, we accelerated investment for growth. In anticipation of demand, we invested $200 million more in CapEx than in the first half of 2011. We invested more in our product offering, including handset subsidies, as we see unmet demand for mobile Internet in our markets. And we continued to invest in people and added staff in our fastest growing categories, with encouraging signs of growth.

I'm also very pleased to report that Paraguay is now our first country to generate more than half its revenues from VAS. This is our aim for the entire Latin American region by 2015, and we are on track to achieve this.

Earlier this week, we announced the acquisition of Cablevision Paraguay, 20 years after we began mobile operations in Paraguay, and four years after we made our first cable acquisition in Latin America. We will now also bring the Tigo brand into the big screen in homes in Paraguay.

In the first half of the year, we returned $350 million to our shareholders through share buybacks and dividends.

And today, we also fine tune our guidance within the parameters of the outlook we previously communicated for 2012 to give more visibility and reflecting our decision to accelerate investment for growth.

Slides 6 and 7, turning to the key financial highlights of H1. Our underlying local currency revenue growth increased slightly in Q2 to 8.9% from 8.4% in Q1. This is mainly the result of increased investment in our new categories. On slide 7, you can see that the acceleration of investment had a dilutive impact on our EBITDA margin in Q2.

Slide 8. The transition from voice to data and from prepaid to postpaid in mobile is diluting margins, not only ours, but also those of most players in our market. Having said that, our margin remains among the highest in the mobile industry.

Slide 10. I would now like to give you a strategic update on where we stand six months into 2012, and nine months since our last Capital Markets Day. I will give you an update on some of our most strategic competitive advantages, our brand, our distribution network, and the progress made in building our categories in H1 2012.

Let's start with our Tigo brand, slide 12. Those of you who have been following Millicom for some time will remember the days when we had a large collection of differently branded cellular businesses. Since 2004, with the first launch of the Tigo brand in Central America, we have succeeded in making Tigo brand -- Tigo our universal brand.

Tigo is now recognized and respected as a brand that stands notably for simplicity and proximity to customers. These attributes are visible in our local taglines, examples of which you can see here. These are all drawn on our brand's origins. The Spanish word contigo, which means with you.

In 2002, our product and service offering has become increasingly diverse and sophisticated. We have introduced sub-brands such as Tigo Money, and we are also introducing a global tagline, Smile, you've got Tigo. On slide 13, you will see an extract of our latest advertising campaign, Smile, you have got Tigo.

Now let me show you a recent TV commercial aired in Latin America.

(Video playing).

Smile, you've got Tigo.

Slide 14. Today, the Tigo brand is one of our most valuable assets. In 2012, it has been ranked number 52 in BrandFinance top 100 most valuable telecom brands with a brand value estimated at $2.7 billion, which is equal to 23% of our market capitalization.

As I've already mentioned, Tigo is also a local brand, and we are pleased to see that in Colombia, only five years since its launch, Tigo is now in the top 10 consumer brands in the country, which we see as a breakthrough given that Colombia is one of the few markets where we do not yet have a leading position.

Slide 15. We have been looking at ways to measure our success in converting awareness of the Tigo brand into customer loyalty. Here you can see an example from one of our Central American markets where over 95% of our customers would recommend Tigo to friends and family. These marketing studies also allow us to take commercial actions to strengthen our brand.

Slide 16 and 17. Most of our success have -- seen in the rollout of VAS in our new categories, has been facilitated by our extensive distribution network. This network enables us to, one, provide our customers with many services beyond airtime reloads, and to build them from these new services.

We have almost 700,000 points of sales across our footprint, with an average of one point of sale for 400 people across our markets. We also deploy direct sales forces in our operations that cross-sell and up-sell Tigo products and services on the streets and doorsteps of consumers' homes.

We have implemented across our distribution network a superior distribution management system which enables us to track and monitor in real time inventories and airtime sales, which some of you have seen in our markets.

Slide 18. As we consider ourselves more as a fast-moving consumer good company than a technology company, innovation is one of the pillars on which we build our growth.

Slide 19. As you can see on slide 19, in Q2, we recorded the highest absolute growth from our four new categories. This revenue outside communication grew year on year by 31% in Q2 from 29% in Q1. At the same time, we managed to develop our communication revenue, growing voice revenue by just under 1%, and SMS revenue by a strong 9%.

Slide 20. In Q2, over 80% of our revenue growth in local currency came from products and services that we launched over the past three/four years.

Slide 21. This slide shows product penetration of our services. As you can see, there is ample room for further growth in penetration and uses of many of our currently successful innovative products.

Slide 22. Through the staffing of our categories, we are investing in future revenue growth opportunities in the form of new product and services. We know we have only six months before our successful products get copied by our competitors. This is why we have to constantly innovate to stay ahead. The entrepreneurial spirit in which we have our roots is still very alive in Millicom.

Slide 23. Mobile data within the Information category is currently our largest single contributor to revenue growth. As you can see on the chart on slide 23, there is a high correlation between revenue growth in mobile data and traffic growth. It is very positive, as it demonstrates that we have the right pricing in place to monetize mobile data. Recently, we have seen an acceleration in traffic growth which we are supporting, given the rapid decline in the production cost of Internet access.

Slide 24. Mobile data penetration has reached 15.5% of our Latin American customer base in Q2. Not only did penetration increase, but so did usage, with an ARPU still up 2% Q on Q.

Slide 25. In mobile data, our strategy is to secure leading positions in our markets, as we have achieved in mobile voice. In transition from voice data to -- and prepaid to postpaid, devices are enablers. They give customers the tools with which to access the Information category. In H1, revenue from mobile data grew by over 44% in local currency. This growth was supported by improved affordability of good quality, branded smartphones, as well as higher subsidies.

At the same time, less than 5% of our total mobile customers are smartphone users. Of our 4.2 million mobile data users in Latin America, we can see a growing use of Android devices, which is a positive development, as Android customers deliver attractive returns.

Slide 26. Given the penetration of our product is being driven by our commercial investment in handset subsidies, our total subsidy cost increased in Q2 by 16.2% year on year. Indeed, we decided to pass on to customers the benefits of the decline in device prices, making access to the category affordable to more customers. Moving customers from prepaid to postpaid not only increases ARPU, but it also reduces churn.

Slide 27. Our Entertainment category brings relevant content to consumers through our mobile and fixed networks. We offer pay TV in Central America, and soon also in Paraguay. We have successfully and continue to grow revenue from ringback tones, notably in Africa this quarter. Revenue in the Entertainment category accelerated significantly in Q2, growing 16% year on year.

In the first half of 2012, we rolled out SMS browsing services to all our customers in Latin America. Rollout in Africa will follow in H2. With this service, we are making access to Internet content available to all customers, even those who cannot access the web with a smartphone.

In certain markets, we have also successfully launched USSD access to Facebook, facilitating access to social networking for more customers.

Slide 28. Solutions was one of our fastest-growing categories in the first half of 2012, and has a pipeline of over 100 planned product launches. Our most successful Solutions products continue to be our zero balance products. Again, in H1, close to 20% of prepaid airtime was borrowed, either from Tigo, or from friends and families. This demonstrates both our marketing capability as well as our ability to find new growth opportunities. We are also starting to generate meaningful revenue from our Tigo Care franchise.

Slide 29. This slide shows our Tigo Care range of products related to health, safety, insurance and welfare. We have rolled out products in this product line in Guatemala, El Salvador, Honduras, Colombia and Ghana. In Guatemala, for example, we now offer insurance for device [death], medical consultation call services, and through our partners, certain types of health insurance coverage. With these products, we have structurally lower gross margins than our airtime lending products, but they increase loyalty, churn reduction, and have very limited CapEx.

Slide 30. MFS continues to develop in Q2, and contributed close to 9% of Group revenues -- Group revenue growth. This contribution is mainly coming from three markets; Tanzania, Paraguay and Rwanda, and for one product, domestic money transfer.

In Tanzania, the penetration of MFS has now reached close to 30% of our customer base. In Paraguay, 19% of our customers were using this service in Q2. We are also offering an international money transfer service in Paraguay in partnership with Western Union, and we expect to launch similar services in other Latin American markets under this partnership in the coming months.

In Rwanda, the growth of MFS services continued to be strong, and by the end of Q2, 12% of our customers were using this service. ARPU for MFS users now stands about $1 in all of our markets.

Today we are offering MFS in seven of our markets, and we expect to launch in Chad, DRC and Bolivia this year.

I will now hand over to Francois-Xavier, who will take you through our financial results for the quarter. Francois.

Thank you, Mikael; and good afternoon, everyone. Slide 32.

In Q2, we recorded local currency revenue growth of 9.4%. Excluding one-off items, revenue growth was 8.9%, up 0.5 percentage points from the growth recorded in Q1. This slight improvement in the growth is a direct consequence of the investments we are making.

The EBITDA margin was 43.4% for the quarter, down 2.4 percentage points from Q2 last year, due to accelerated investment in our categories, as Mikael outlined earlier.

We invested $247 million, or 20.9% of revenues in CapEx during the quarter, of which $32 million related to spectrum acquisition and license expansion in DRC.

As I mentioned last quarter, we are attempting to spread our investments more evenly throughout the quarter -- throughout the year. The cash generation in the quarter was negatively impacted essentially by the timing of investments and tax payments.

Now let's look at the performance in each of our three regions, starting with slide 32 with Central America -- 33, sorry.

Revenue from mobile and cable operations in Central America was up by 7.9% year on year in local currency. Adjusting for the one-offs in Guatemala last year, local currency revenue grew by 6.2%, increasing by 1.5 percentage points over Q1 2012.

Solid growth in Honduras and Guatemala once again more than offset the top line decline in El Salvador. We experienced a 3.4% year-on-year decline in mobile ARPU in the quarter.

Our EBITDA margin was 49.7% in Q2, declining 1.9 percentage points from Q2 2011.

We have accelerated our network investments and increased our handset subsidies relative to the prior year, as we see clear opportunities to develop the information on MFS categories further in Central America. Voice pricing pressure in El Salvador also contributed somewhat to the margin decline in the region.

Slide 34. In South America, revenues increased by 13.5% in local currency on an underlying basis, and all three markets reported a strong performance.

We added close to 210,000 new customers in the quarter, more than 80% of which were in Colombia. Mobile ARPU was up by 2% in local currency as a consequence of our ongoing focus on mobile data and other VAS. This also reflects our success in cross-selling and up-selling services to existing customers, as well as the addition of high quality, new customers to the Tigo network.

In Q2, around 18% of our recurring revenue in South America was generated in the information category, the highest level in the Group.

Increased subsidies and other investments to grow better market share contributed to a year-on-year decrease in the EBITDA margin by 3 percentage points to 39.8%. The margin was also impacted by a new tax on revenues in Bolivia implemented in 2012.

We will continue to encourage and support our customer transition from voice to data and from prepaid to postpaid in the region, especially in Colombia, by offering quality services attractively bundled with smartphone subsidies.

Slide 35. Now let's turn to Africa. Revenue in Africa grew by 5.7% in local currency in the second quarter. Revenue was negatively impacted by the strengthening of the US dollar against many of the currencies in our African markets.

Performance in our African footprint remained mixed, with some operations performing strongly, such as Tanzania and Rwanda. We expect increasing competition with the arrival of new players in both Ghana and Rwanda.

Despite this, we increased our net customer additions quarter on quarter, and saw a clear slowdown in local currency ARPU erosion by 3.2 percentage points.

The EBITDA margin for the quarter was 38%, down 2.4 percentage points, as we focused on our affordability perception. We also recorded a provision for bad debt of $5 million on international traffic in Senegal.

CapEx in Africa in the quarter was $38 million, higher than in Q2 2011 and included investment in spectrum, in DRC, in the attractive 900 megahertz band, as well as an extension of our existing license from 2017 to 2024. The additional spectrum will enable us to expand our network more efficiently in this vast and populous country.

Slide 36. Normalized EPS decreased by 2% year on year to $1.74.

Slide 37. Our tax rate for the quarter was impacted by higher withholding tax on upstream cash. Despite the fact that corporate tax rates are increasing in a number of countries, we are confident that we will able to maintain an effective tax rate of less than 30% for 2012.

Slide 38. Our free cash flow for the quarter was $98 million, reflecting substantially higher CapEx this quarter than in Q2 last year, as we aim to spread our network investments more evenly between quarters.

Slide 39. In Q2, our net debt to EBITDA ratio increased by 0.8 times, as we returned $350 million of cash to shareholders in the form of both dividends and share buybacks.

Slide 40. Turning to our debt maturity, the average maturity of our gross debt now stands at three years. During the quarter, we issued a $180 million non-recourse local currency bond in Bolivia at 4.75%, with an average maturity of five years. 52% of our gross debt is at fixed rates, reducing our exposure to interest rate volatility.

Slide 41. Our recent bond issuance in Bolivia has further diversified our sources of funds, with bonds and development finance institution debt now accounting for 39% of our total debt. We aim to be less dependent on the traditional bank financing over time.

Slide 42. Here you can see the split of our debt by currency and interest rates. At the end of Q2 2012, 59% of our debt was not exposed to currency fluctuations. Our debt structure has been designed to provide some natural hedge against both currency and interest rate risks.

Slide 43. You can see that of our total debt of approximately $2.5 billion, 68% is non-recourse. It is also our strategy to limit corporate guarantees as much as possible.

Slide 44. We were pleased with the agreement reached to acquire Cablevision Paraguay, the leading player in cable TV in the capital city, Asuncion. The acquisition is subject to regulatory approval. This acquisition will enable us to further accelerate our growth in fixed broadband, and to improve the quality of services to customers through network integration.

Slide 45. I would now like to comment briefly on shareholder remuneration. In June, we paid a $2.40 per share ordinary dividend to shareholders. In Q2, we bought $106 million of shares in line with our previously communicated program of up to $300 million. In the absence of further external growth opportunities, we remain committed to returning excess cash to shareholders.

Slide 46. In 2012, we again aimed to strike the right balance between top line growth, profitability, cash flow generation, and return on invested capital. We expect the full-year EBITDA margin to be around 43%, and the operating free cash flow margin to be around 20%. We expect CapEx in 2012 to increase, but to remain below 20% of sales, excluding spectrum acquisitions, as we invest in better capacity, IT and billing platforms.

That concludes our presentation. We will now be pleased to respond to your questions.

We will start with questions from the audience present here today. If you can please introduce yourselves, stating your company name.

J.P. Davids, Barclays. Two questions please. The first one is on distribution, which you have mentioned is a strategic advantage or key pillar to the strategy. How does the distribution model change, or if at all, as you move from the voice to the data world?

The second question is on the margins, and you've clearly communicated a reinvestment in this growth. I guess it seems there is clearly an incremental step-up from your side that's different from what you saw at the beginning of the year. You've mentioned the areas. Maybe you can mention some of the countries where you're seeing this extra demand and maybe the payback period that you foresee for this incremental investment.

Thank you.

Okay, let me start with the distribution side. With the increased amount of smartphones or data phones out there, what we are in the process of doing right now is out of those 700,000 distribution points, which are the ones that service our higher ARPU customers that tend to also be the smartphone users.

And then what we can do is train -- educate and train the people who run these retail shops to sell our packages and products. Rather than depending on that the customer walks in knowing what they want, somebody can educate them based on their calling pattern, for example. That's an important thing we do there.

The other is MSF, of course, where we take the largest of the 700,000 retail points, ensure they have enough liquidity, and re-brand them as Tigo cash in and cash out points.

So it's a continuous redistribution, so you continuously evolve and innovate and move forward.

Regarding the margin and the additional investment, if I understand properly your question, was it -- is it more specific to a given region? We've said that it is across all regions; maybe a little bit more in Latin America because we are investing much more in subsidies, and the debt opportunity is a little bit larger, at least for the time being in Latin America. Also, we are starting the investment as well in Africa, but we don't have yet a 3G license in all countries.

The feedback is actually very good. We disclosed that a year ago, nine months ago in London during our Investor Day. We had -- nine months ago, we had a return on invested capital on debt which was at that time around 17% for Latin America that we compared with [the WAC] at 11%.

And since then it has even improved, so I think that we are getting closer to 20% return on invested capital as far as data is concerned, which is largely the consequence of the slide that you saw before, with the fact that we have a parallel curve between the growth of traffic and the growth of 3G revenues, which is probably not that common in the industry.

And let me give you another example that happens to be a smaller category. Take the Tigo Care, for example; you know, either insurance or a doctor's appointment, and so on. These are categories where we have a lower gross margin because we tend to operate with a third party. Last quarter, we generated about $10 million in revenues in this category from a virtually zero base.

And this is a category that's very limited CapEx, limited marketing investment, lower gross margin, but is great for customer loyalty and churn reduction.

Barry Zeitoune, Berenberg. When I look at slide 20 and the mix of revenue growth, it's pretty apparent that this quarter, like other quarters, we're seeing lower growth coming from voice and higher growth coming from data, which is making up the difference.

I'm just wondering how, when we think back to the Capital Markets Day and we had the original targets for growth, how does that change our medium-term view on margins; the fact that we've got more growth coming through from data than is now coming through from voice?

And my second question is really on the slide where we showed the margin progression and we saw 0.6 percentage points of impact from the one-off items. I just want to check; are these completely one-off items, or are these new taxes that are going to be repeating in quarters ahead?

And then finally, you mentioned that the Cablevision acquisition is going to be financed by debt. Is your medium-term view of where you expect your leverage to be changed as a result of the acquisition?

Maybe I take the last question. This acquisition of Cablevision is important because it shows you what we want to do, which is to make acquisitions for which we'll get a good return and that are complementary to what we do, with a good return with leading position, which is the case of this acquisition.

It is not a big acquisition. It's $150 million, so we don't want even to change our shareholder remuneration of the consequences. It's an opportunity to marginally increase our leverage. We are talking only of $150 million.

So this would be obviously different if we were going to do larger acquisitions in the future than obviously what I'm saying for Cablevision might not be valid in that case.

Talking to the -- about the one-off items, a 0.7 percentage point, it is essentially -- it is a pure one-off which is essentially linked to this write-off that we had to do on international traffic in Senegal. So this is not a recurring item.

On the margin point versus the Capital Days, I think we're seeing more persistent pricing pressure in some markets in Africa than we anticipated. And also, El Salvador has been a market where there is quite significant pricing pressure. At the same time though, we still have a very strong growth on SMS. You can see that's going at 9%. And that's virtually at 100% gross margin, so --

(Inaudible - microphone inaccessible) ...we expect margins to be different though. Or a year ago, we were talking about mid 40s. I know for 2012, we're talking about 43% now, but is our medium-term view changed at all, or not really?

We have said for two years now that structurally anyway, we will go into a lower margin going forward due to the fact that we are entering and developing and growing in categories that have structurally a lower margin such as Mobile Financial Services, Solutions and Entertainment.

We said that we were not especially concerned by this underlying trend due to the fact that this -- it happened that these three categories have lower EBITDA margin, but they are not really CapEx intensive.

Just to give you some indication, so far for Mobile Financial Services, we invested a little bit more than $1 million per country. If I were comparing, for example, the last [license] that we got in Rwanda, the amount of money that we invested in Rwanda compared to, for example, what we generated in terms of MFS in the neighboring country, in Tanzania, frankly speaking, I would rather go for the low CapEx. Even if the margin is lower, I would rather go for the lower EBITDA margin but low CapEx business than the traditional one with much more CapEx.

So structurally, the margin would decrease. Will it decrease further for 2013 and the following years? We are still working on our planning for next year so we cannot comment on that. But it is sure that there is an underlying trend.

Also, I have just one point there. If everything else been equal, you also have different growth rates between the higher margins, Central American regions, and the lower margin Africa and South America. So the mix is impacted by that.

In the long run, that's quite a bullish message on CapEx to sales then, and where we might expect --?

I think over the time, I think the new categories just have to grow in scale, and I wouldn't expect that that leverage would come in place in 2013. I think it's more 2014 and beyond that we could start to see that [impact].

Especially that you know even this year and again next year, we still have a fairly high amount of CapEx coming from IT, because we said that we had about $300, more than $300 million of CapEx to do over three years, 2011, '12 and '13, and the bulk of it would actually come in 2012 and 2013.

But as Mikael said, probably from 2014 onwards, due to the fact that a significant part of the growth will come from non-CapEx intensive categories, yes, we should see that trend happening.

Laurie Fitzjohn, Citigroup. Firstly, given your decision to invest more in subsidies than value-added services, do you consequently expect higher growth, or should we expect growth towards the upper bound this year, and also into '13 on the organic on the 8% to 10%?

Secondly, in terms of DRC and Senegal, interesting to note that you flagged these in Q1 as higher competition, but they seemed to have had pretty decent net adds in Q2. I wonder, are you seeing the competition start to rationalize slightly in those markets?

And then lastly on El Salvador, it's been tough for a while, but is Telefonica starting to act a bit more rationally, given it's a while since AMX and Digicel merged?

With Telefonica and El Salvador, the answer is, no. Let me -- DRC is a slightly complex environment because a minimum pricing is set, and basically, therefore, it's very difficult to come out with product plans that would give you an edge. So there's very limited room what we can do there. So we believe that the Mobile Financial Services product that we will release in the next few months would really give us an edge in that market.

And in Senegal I think, yes, it's been a little bit more competitive, but remember, this is the market where we are in this dispute with the Government, so we are slightly behind in investing what would be required to capture the opportunity there.

And regarding the growth, I think there are two parts to the growth. There is the innovation part on which we are very confident, and I think it's working very well. As you can see in this quarter, in quarter 2, actually, we did the best -- we got the best contribution from the growth from innovation that we ever had, so which mean that it is really working as planned.

What are the risks? Probably some uncertainty on the communication category, especially on pricing, which is something that we do not always control, although we participate, we may participate in it, which is the pricing level and the pricing pressure that we may have on voice, which is something that we suffered a little bit lately. That could be factor that could influence potentially the growth in the future.

We are pleased to see that in Q1 and Q2 again we had the resilience voice business, and MFS growing at 9%.

And if you remember, I pointed out that in Paraguay, now 50% of the business is non-voice. So if you keep on accelerating this non-voice part, you are less and less sensitive to any pricing movements on the voice side.

Okay. Thank you.

Mark Walker, Goldman Sachs. First question on South America. I get in the release where you talk about subsidy investment depressing your margins in that region, but can you talk a bit more about why the revenue and ARPU growth slowed?

And then secondly in Africa and just a follow-on from Laurie's question, you mentioned that you -- it's your expectation that growth should improve in the second half because of some initiatives that you're going to implement. Is that really just the MFS in Ghana -- or, sorry, in the DRC and Chad that you're talking about?

And then finally, more general question on subsidies. If subsidies are increasing 16% year on year, how do you expect that to trend through a medium-term view?

Thank you.

Let me start by Africa. I would be somewhat cautious on the growth acceleration in Africa in the near quarters. What we are doing is putting together a team that are going to be able to support an accelerated voice growth in Africa. There's lots of opportunities around that; simple services such including Mobile Financial Services. Putting all that in place and having an ability to impact the market will take some time.

In North America, you are talking about the revenue and the ARPU growth slowdown. Actually, the ARPU kept on increasing. The ARPU in South America has been increasing for almost the last two years I think now, constantly quarter over quarter, which is very positive.

There has been a little bit of a slowdown in the growth, which is partially linked to the fact that we see the economic momentum in South America slowing down somewhat. You probably noticed that in Brazil, for example, the economic growth has -- is much lower than it used to be, which probably -- which impacts somewhat our growth in -- especially in Paraguay and probably in the other countries as well. But the growth momentum remains strong.

You want to comment on the subsidy trends?

Yes. In subsidies, you basically have falling device prices coupled with -- and then we play around with subsidies to -- with rate plans and paybacks. So our hope is that the device prices will continue to come down. I think we are -- right now a good smartphone will be somewhere at $250 for a customer; a medium at around $50, and low end below $100.

And naturally, if over time these prices come down, we could either reduce the amount of subsidy we give per phone, or we could just accelerate the amount of phones we put out there in the market.

Soomit Datta, New Street Research. A couple of questions, please, first on the cash return. You said you'll look to return excess cash towards the end of the year. Can you give any more specifics on when we can hear anything?

And you may not want to comment on this, but ballpark, you returned about $1 billion last year. You've signaled about $550 million so far through buyback and dividend; $150 million for Cablevision in Paraguay. Would it be wrong to think that potentially there is $300 million which we can hear about later in the year?

Secondly, just on Africa, I think the initiatives you've talked about are focused around data, but I noticed the voice revenue trends in local currency slipped, so they're down minus 1.5%. No talk about turning that around though. Is that basically what we can expect going forward from African voice?

Then just finally on Colombia, I think there's mention in the press release about voice ARPU under pressure. Can you talk a bit about where that's coming from, please?

Thanks.

Yes, let me start with Africa voice. No, we haven't given up the plan nor the ability to turn that around. That's very specific to certain markets. There are many variables you can play around to address that one.

In terms of any more shareholder returns, we don't have any new announcement on that point today.

It is true that we returned $1 billion in the last two years, but we stick to what we said before that we will return excess cash, okay? We don't want to commit any new figure at this stage. We have said as well that this might be impacted, although we confirm that it is not the case for Cablevision Paraguay, but it might be impacted by any external growth opportunity if it happens. If it doesn't happen, you saw what we did in the last two years.

Do you want to comment on the Colombia voice?

It's a seasonal thing. I don't think there is anything specific to read into that. Colombia voice, yes.

Peter-Kurt Nielsen, Cheuvreux. If I may just return to the subsidies question then. How much of these -- of the handsets do you currently subsidize? Is it possible to be a bit more specific on this? And how long are you normally tying up your customers?

And secondly, on the move you talked about earlier, Mikael, towards postpaids from prepaids. How do you see the scope in the overall subscriber base of actually a bigger share here of moving towards postpaid model, and how does that impact your own business model overall?

Thank you.

Let me start with the postpaid, or prepaid to postpaid move there. We are not fighting any trends. We welcome any trends. We just have to figure out how do you take advantage of that.

Normally, when you move customer from prepaid to postpaid with a data plan, you get a significant uptick on ARPU. And you take something in Colombia, roughly half of our revenues are already postpaid as an indication. And this is the most mature market, of course, we have in terms of GDP per capita, so that's natural that should happen there.

And in fact, in Colombia, our data market share is twice our voice market share, so we are winning in the space. That's very important for the future.

In terms of subsidy levels and so on, we don't for competitive reasons want to be too explicit about that point. If we have -- paybacks are normally six to eight months. We have some markets the length of the contract is set by regulation, so we have anything from 12 months to 15 months to 18 months. Some high value customers that really give a top end smartphone, we can have 24 months contract, for example. But we are aiming for a payback let's say on average on six months.

Anyone have some questions on the phone? Can we take some questions from the phone, please? Operator?

Andreas Joelsson, SEB.

Just wondering if you could explain the differences in ARPU trends in South respectively, Central America? It seems Central America having declining ARPU and we have increasing ARPU in South America, and just wondered why that is.

And secondly, if you can give us a definition of excess cash; what kind of leverage do you have when you think you have excess cash?

Let me start with the ARPU trend between South and Central. You have a relatively subdued economic growth in Central America. That's the one market area which is very dependent on Central Americans migrating to the US and sending money back, and the remittance side. And we also have more price pressure in El Salvador which is basically dragging down a little bit the ARPUs in the combined.

South America has a stronger economic growth. Colombia's doing well. Bolivia's doing well. And Paraguay's doing reasonable well, although it's been slowed down now with the economic slowdown in Brazil.

So you have a combination, basically, some macro factors that are -- and specific placing actions that are making the delta.

I'm not going to give you a precise definition of excess cash, but we were in terms of net debt to EBITDA ratio at 0.6 times at the end of Q1. We were at the same level last year. We are 0.8 times today because we just paid the dividends and did some share buyback

What we have always said is that we would feel comfortable with a net debt to EBITDA ratio is 1 times, so we are getting closer to that point, given that we are even going to get a little bit closer with the acquisition of Cablevision Paraguay, because we said that we are going to finance it through additional debt.

And so from that point of view, you saw what we did in the last two years, so we returned what we define as excess cash. But I cannot -- once again, we don't want to give any precise amount at this stage.

Okay, thanks.

We have no intention to be sitting on piles of cash on which, as you know, we have not such an attractive return. So we value a lot shareholder remuneration, so which is the reason why we will not hesitate as we did in the last two years to return what we believe is excess cash.

Perfect. Thanks.

Operator, can we have the next questions, please?

Rizwan Ali, Deutsche Bank.

My question is the pricing pressure in Africa. Last time, or in the last conference call, you said that there was a lot of off-net pricing decreases. Does that continue to be the case? And is it again Barclay which is being more aggressive in Africa? Can you elaborate on that?

And second is postpaid. Obviously, especially in Latin America, there is a big trend towards postpaid phones. Where do you see as percentage of your total subs postpaid going or representing? Will it be closer to 20% of your sub-base eventually?

Okay, let me start with that one. If you take Latin America, I think you have roughly around 30% of the customer base with an ARPU greater than $10. I think that would be a natural penetration target across Latin America.

In terms of the pricing environment in Africa, yes, a lot of the pricing moves was off-net, and as you all know, the elasticity on off-net, i.e., when you go from one network to another one is very limited, because the behavior in the mobile industry is that you have three or four people on your network that you call a lot; so when you reduce that price, you tend to have great elasticity. People are thankful and just talk more, and there is a general perception that it is expensive to call to other networks. So even if that tariff falls, if you are an ordinary household, you might not have so many people to call so it's -- in our books, it's revenue loss.

But a lot of the moves were driven like that, I think. Yes, it was both Barclay and Vodafone who did these moves. And so we're still suffering through that the lack of those higher across-net revenues we had.

Maybe one comment on postpaid, because we don't see postpaid, the move from prepaid to postpaid as something negative, because it brings additional or increased ARPU to start with. It increases loyalty and registry churn. And it's not always negative in terms of EBITDA either, because it is true that we have the cost of the subsidy but we save on commission for reloads as well. So net-net, it's not something that we see as negative.

But on average, in at least Lat Am, we find that postpaid margins are slightly below prepaid margins.

In terms of multi (inaudible), but on a higher ARPU. So in absolute value in terms of EBITA margin, it's not always negative.

You're right, but in terms of trying to forecast trend in margin, as you go more and more towards postpaid, should we then expect continuously or [potential] --?

Yes, I think you will see more incremental EBITDA margin, but maybe some margin reduction from that one.

Now on the other hand, we've heard that data margins tend to be much higher than voice margins, so I'm a bit surprised when you said that in your case, you find that as you have more data, margins may come under pressure.

Can you repeat that?

On average, we've heard from operators that data margins are much higher than voice margins. So as you have more and more value-added services and data services on your network, your margin ought to improve.

Yes, but the global margin is not only linked to data to start with, because as we said, we have the dilution from the other categories, namely Entertainment solution on MFS. And the other side of the equation is the fact that when we say we have a better margin on data than on voice, it is at gross margin level. So at EBITDA margin, okay, you need to take into consideration the impact of subsidies as well.

But also, there is not much of a difference. It's not significantly lower in terms of EBITDA margin on data versus voice.

We tend to probably have higher voice margin than the developed world because we have more on-net calls which have the higher margin versus the across-net calls. I think people in the developed markets are less sensitive if they call other networks.

We are going to take probably one more question from the phone and then move back to the audience. Operator, can we have next question, please?

Thomas Heath, Handelsbanken.

Two questions, if I may. Firstly, you show the charge there on data traffic and revenue growth, or data revenue growth being correlated. Is it 1 to 1 in terms of just to make sure they get the scale right; that sort of a 1 factor increase in data traffic leads to revenue, data revenue growing 1 times as well? Just to check that. And also, do you consider that sustainable? Because it truly is an outlier compared to other operators.

Secondly, I just wanted to check on El Salvador, the price pressure there. What would it take to change that and what's the risk that we see a similar development in other neighboring markets?

Thanks.

Yes, let me comment on El Salvador and that's maybe a bit of a speculation there. You know that there is this intent to try to merge Digicel and America Movil, and that's been frustrated by the lack of regulatory approval, and that has probably driven some of that pricing pressure we have in that market. We haven't seen a similar trend in the other Central America markets.

On the slide on data traffic, I confirm that the scale is the same, so which means that the parallel curve that you see between traffic and revenues on data is a comparable base.

You probably noticed though that in the last two quarters, we had a little bit more growth on traffic than on revenue, which is always something that we were monitoring. We are a little bit less concerned than we were, because on the other hand, we have a very significant decrease in the cost of production of data, megabytes and so forth.

So we monitor the relationship between traffic and revenues on data as well as the cost of producing data, with the objective of maintaining a gross margin that is consistent, and consistent or even why not improving over time.

So -- but we want to make sure that we don't have the same kind of trend that we see in the developed world where you have massive growth of traffic and a smooth growth on revenue. So we are not there, and we are still with a fairly parallel curve, but it may divert slightly or marginally over time.

Okay, thank you. That's very clear.

Can you just remind us when we're going to see an impact of the MTR cuts in Colombia? Have we seen anything in Q2 in there and can we quantify it?

I think it's too early to see any trend. It went through in --

It went through in April.

It went through in April.

So we should be seeing the evidence of it in the Q2 number then?

Yes, but bear in mind that it is a [certain] small decline in the base coming from a low base already.

Okay.

It's a 50% decrease over three years though, three years. So don't forget that we had a 50% decrease I think in 2008, but it's a 50% decrease over three years off a base that has already been reduced by 50%, and we are far, far less dependent on voice today because our data business has really developed extremely well and, as Mikael said, we have a market share on data which is about twice as high as the market share that we have on voice. So we consider that -- not that it is a non-issue, but it's not a very significant issue for us.

But that would be one of the reasons why voice growth in Colombia was a touch slower?

Marginally.

Marginally.

Yes.

And it wouldn't have impacted the margins very much?

Not significantly.

Okay. And then also in Colombia with the auction coming up, I've penciled in about $150 million. Am I being very aggressive or very conservative?

Well, we don't want to comment on what we will be prepared to pay for any spectrum when and if it comes for auction.

Okay.

The only thing we can comment is that the auction was supposed to be issued in June and it has been postponed I think to October or November, and so the process seems to be a -- take a little bit more time than expected initially. but we'll see the way it goes.

Question on your capital structure. With 60% of your debt now locally held, are there other opportunities to push debt down to local levels, or is that where it's going to sit for a while?

Well, if we could have 100% of our debts locally in local currency of -- with reasonable maturity, we would do it. It's always a trade-off between costs and risk as well, because in some markets, we could take even more in local currency, but it is usually more expensive as well.

It's limited by another factor, which is the fact that there is not always the level of liquidity or the level of debt in some of our emerging markets, so long-term debt in local currency. So it's either not available, or either too expensive.

If you take -- let me take an example. In Honduras, for example, we have -- there is a market for medium-term debt in lempira, the local currency of Honduras, but it is, I think, something like 50% -- at least 50% more expensive than dollar. So in that case, for example, we have decided to take half of our debt in US dollar and half of our debt in local currency. So it's always a trade-off between the cost and the risk.

Okay, thank you.

Operator, can we take some more questions from the phone, please?

Ric Prentiss, Raymond James.

A couple of questions. I want to come back to something you mentioned about Colombia. You had mentioned that your share of data is twice what your share of voice is. Can you update us on what those shares are and how you achieved them?

And then my second question is at the analysts' day back at September, you had an interesting slide talking about how you had moved people up to smartphones or to 3G phones and seen a large uptick in their revenue over the first year. Can you update us on what you've seen as far as when customers do upgrade to 3G or smartphones, and what the ARPU trends are?

Yes, let me start with the last question. We are -- from a commercial point of view, we are a little bit holding back too much disclosure on how much more ARPU we are getting from every prepaid customer who moves to a postpaid data plan, or a postpaid feature phone customer who moves to a postpaid smartphone plan. But we still clearly see significant increases in ARPU when that happens, so that is a very clear trend.

Yes, the data is actually public in Colombia. It's the regulatory authority there produces it I think four times a year. And I think approximately, we have a 27% share of the data market and around 12% share of the voice market.

Yes, so double that. And then as far as the margins go on the data, are you still seeing better margins because you don't have to pay the termination rates? I know you had the earlier question about the other value-added services, but it seems like data would have by itself better margins because of the lack of termination rates.

Yes, data margins in Colombia are quite good compared to our voice margin, because us being the third operator, we still have a proportionately higher part of our calls across net. So if you just compare voice gross margin versus data margin, at this time in Colombia, the data margins are a bit more attractive.

And how are you achieving that higher share on data than on voice? Is it distribution, or what explicitly are you doing to really create [those volumes]?

I don't know if I want to disclose all our secrets, but let's put it this way. We have a lot of feet on the street. We have a lot of young people with ponytails and tee-shirts who are out there demonstrating to customers how to use data services, because most people really need to be taken through a demonstration so that they understand how to use this, unless you have a set of very smart teenagers at home, particularly if you are an older customer. So it's really about having feet on the street.

And then naturally having plans that matches the consumer behavior, the consumer needs. Our most popular data plan is what we call our young adults social plan. So it's typically on a BlackBerry; it's unlimited access to social network, but not YouTube, emails and BlackBerry Messenger services, and then we have a voice pack on top of it. That will cost around $25 in Colombia. That's like a -- that's fitting a product that matches what our customers want. So insight in the product offering and feet on the street in its simplicity.

Great. Thank you.

Erik Pers, Danske Markets.

I just have a big picture question. It seems to me that now that you're undergoing this structural margin decline, and with your revenue base being -- approaching $5 billion, it might be difficult to accelerate the revenue growth significantly. It seems to me that you inevitably are in a period, potentially quite long period of slow EPS growth. How do you look upon that from a management perspective? Is that just an inevitable transition period that they have to go through, or can you do something maybe on your P&L lines below the EBITDA line or something else to find a way around that?

The other question I had was just on margins on new services. Is it -- the fact that they have a lower margin that's been partly discussed already, but is it an effect -- to what extent is it an effect of these services having a smaller scale and needed to --? Once you gain scale, will it be possible that you have a very high margin of them eventually? Or do they actually -- are you confident that they actually in a mature state will have a lower margin than voice?

Thank you.

Okay, let me start in the margin end. Yes, in the beginning, they are lower scales. You take something like MFS when the early part of the MFS margin is basically sending money from around the country, but the second stage, if you start doing peer-to-peer transactions without taking out the money, then the margins will go up over time.

But I think there are structural differences. On the MFS side, for example, we share the distribution margin with the retail trade that offers the service. So let's say, on average, cost to send money would be 4%. We record only -- we record the 4%, but we pay half back to the retail trade. So basically, we would have a 50% gross margin when on voice you are up at 85%. So naturally, that margin could expand over time with other services with higher margins.

The Entertainment category is typically about sharing third party content, so you are more of a revenue share basis there, and so that would also, therefore, have lower margins.

The Solutions category, the lend -- the zero balance product where we lend products, either we lend airtime to our customers or there is a transfer between customer to customer, have very high margins.

But over time, some of these categories, like the MFS, could drive higher margins, but all of these are less CapEx intense, so the mix, the revenue growth we get out of this over time will have a very attractive ROIC.

And reflecting on what Mikael just said on your -- to answer your more specific question, you seem to imply that we will have a lower growth in the future. We don't start from -- with that assumption. We believe that we have large growth opportunities, which is the reason why, by the way, we have decided to accelerate our investment for future growth. And it happened that in Q2, we had a higher growth than in Q1. So we don't see ourselves ex-growth. We believe that growth is part of our DNA.

It is true that margin-wise, we have said that for some time, and it has happened last year and it is happening again this year, there is an underlying trend to reduce the EBITDA margin. But as Mikael just said, we look at -- it doesn't mean that we don't look at EBITDA any more, but we look more and more return on invested capital due to the fact that large part of the growth today, and it's probably going to be a larger part in the future of the revenue growth that we have, is coming from categories that have hardly any CapEx.

So there might be a little bit of a transition period during which we may have more -- still a little bit of CapEx. For example, we are catching up, as we said earlier, in terms of IT, but directionally, we think that it's more relevant to look, more and more relevant to look at return on invested capital as well as EBIT.

That doesn't mean that we are giving up on EBITDA, but once again, you know for example, when you take a business like Entertainment, obviously, this is -- a large part of it is more linked to buying content and reselling it, so no CapEx but lower margin.

Okay. Excellent. Thanks.

Lena Osterberg, Carnegie.

I was wondering a little bit on future M&A targets, because now you have managed to secure cable assets in most of your Latin American countries like Bolivia and Colombia. So I was wondering, is there anything that's suitable in size for you in those markets?

And also, I was wondering on Cablevision on the acquisition, what do you believe is the longer-term margin potential for that? You managed to lift Amnet's margins quite significantly after the acquisition, and also, if you could say something about if there will be any integration costs in the second half of the year for that acquisition?

And then also, I was wondering on the ruling in Senegal, when you expect that to come.

Let me start at the last point. I think the ruling is expected any time between now and, I would say, the end of the year. I understand that some of these arbitration judges are probably on summer holidays now, but I think it should be within the next three or four months it should clearly come out there.

So regarding the integration costs for Cablevision, but first of all, Cablevision is a limited -- it's a fairly small acquisition. We are talking of a business of about $50 million in revenue, so there might be some integration costs, but you won't see it. So it's not significant. My view is that it will be covered largely by synergies as well.

Do we have --? You were talking about the margins on cable. Actually, if we look at what we did with Amnet, which is the reason why we did that acquisition of Cablevision, Amnet that we acquired almost four years ago, we have been extremely happy with what we have achieved.

We had a growth in revenues which was at about 13% on average compounded annual growth over the last four years, so which was -- when, at the same time, we had low single-digit growth in revenues in our mobile business in the same countries.

So it gives you an idea of the potential, and at the same time, we increased significantly the EBITDA margin. And it happened that the EBITDA margin of the former Amnet business, which was largely rebranded to Tigo, by the way, that the margin on this Amnet business is actually very close to the average margin of Millicom. So we don't see that as a drag on our overall margin.

Do we see other opportunities in cable assets? We keep on monitoring the situation on a country-by-country basis. It is true that now we have covered Central America and we have covered Paraguay. There are some remaining countries in Latin America, so we keep on monitoring the situation.

In Africa, there is not much at this stage, but things could change over time. But that doesn't prevent us as well from developing some of -- some part of this business on an organic basis, which is something that we could do as well, which we started to do in Paraguay and before we did the acquisition of Cablevision.

And in Colombia, are those companies -- think a bit big scale-wise for you, or are they what you would consider a good size acquisition?

I wouldn't comment any more beyond Francois' comments there. It's a moving field. Are we going to buy out America Movil in Colombia? Probably not, no.

I was talking about the cable companies, not NAMAX.

Yes, we don't have a cable presence in Columbia today, and that's something that we might look into the future.

Thank you.

Welcome.

Kevin Roe, Roe Equity Research.

A couple of questions. First, Mikael, could you update us on any greenfield opportunities out there? And also in-market consolidation, an update there would be helpful, and maybe some of the more stubborn markets like Ghana.

And a second question on -- well, actually, I'll start there and follow up.

Yes. There are not that many greenfield opportunities left in this world, so I think an expansion for us would probably be buying out the number one or number two operator in the market, and then apply our skills on that one from categories to distribution. Or potentially a third operator, if we could see a path for that operator to get to that number one or number two position.

In-market consolidation makes sense. I think fixed operators in Ghana are too many. At some point of time, there has to be a rationalization, and we are not yet there from a decision point.

Okay. And lastly, for the second half of the year, are you monitoring any new developments in regulatory issues or tax issues that could impact your business?

Thanks.

Yes, we are, of course, monitoring this, but I don't think at this stage we have any visibility of any significant changes that would have an impact on the outlook we've given for the year.

Very good. Thank you.

Thank you.

Rodrigo Villanueva, Merrill Lynch.

I was wondering if you considered that a portion of your network CapEx is related to your new categories or not.

And also, I was wondering if you could provide a margin breakdown for your revenue categories.

Thank you.

On the revenue breakdown by -- on the margin breakdown by categories, for commercial reasons, we are not providing that information.

Also, we provided some ideas of the one related to the other last year in our Capital Market Day, so you can maybe have a look at it. I think that they are on our website.

And you were talking about the network cost, is there a link to the new categories. As we said, the new categories, with the exception of Information, 3G is not very capital intensive. It doesn't mean that there is nothing, because there is a little bit of IT capability, especially on the billing side and the CRM side. But the rest of it -- and, obviously on 3G, we have a significant amount of CapEx. But as we said earlier, we are very happy on the return on invested capital that we get on that part of the CapEx.

Understood. Thank you very much.

Bill Miller, J.M. Hartwell.

How many shares did you actually repurchase for the $106 million?

And the second question would be have you done anything with more of your towers? Are they still being up for sale, or is there another company that is going to enter the fray?

And third thing is do you have trade subsidies for your handsets, and go from post to pre, what impact does that have on you? When you have more handset subsidies, in other words, do people tend to go from pre to postpaid?

On the number of shares, I don't know the exact number. It's in the press release, but I think it's around 1.2 million shares that we bought in Q2.

On the towers, we still transfer a few towers to the tower operators we partner with; Helios in Africa and American Towers in Columbia. We transferred some towers in the course of Q2, and we would still transfer some remaining towers in H2 2012 as well as at the beginning of next year.

We are still looking at options to extend further these tower deals to other countries. There might be some [side] opportunities maybe next year. We believe that we have done the bulk of what could be done, but there might be some side opportunities.

I didn't get your question on the subsidies.

Yes, before I go to the subsidies there, I just want to point out that we said many times publicly that we are also very interested in sharing a full network, [a la] Tele2, for example, in Sweden, that shares a full network on 3G with one competitor and with LTE on a second. We think that would be a very good model, particularly when you have new things like LTE coming into the marketplace.

So interested in full network share; I think it's a great way of driving down the costs and just compete on consumer insights and product and services and distribution.

In terms of when we go to pre to postpaid, and when we move from smartphones to -- feature phones to smartphones, we have a significant increase in ARPU. And we showed nine months ago in the Capital Day in London some examples of that. We're a little bit careful with that information at this stage for competitive reasons, but there's still very good growth, and that's why we're actively pushing for that.

Great, thanks.

Thank you. There appear to be no further questions.

Thank you very much. Is there any further questions from the room here? No, I don't think so. Thank you. So, Mikael, would you like to conclude maybe on today's presentation?

Yes. It's been a somewhat challenging environment, and so we are particularly pleased that we really started to harvest the benefits of our investment in improving our customer proposition.

As we said, we are increasingly focused on innovation and sustainable innovation. The new organization structure will allow us to drive that at an increasing pace going forward.

So I'd just like to thank all of you who joined us today, both people here in London and then in the ether space there. So thank you very much and hope to see you soon again.

Thank you.