Millicom International Cellular SA (TIGO) 2007 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to today's Millicom 2007 Second Quarter Results Conference Call. For your information, this conference is being recorded. May I also remind you that this call is being audio-streamed over the Web and is accessible at www.millicom.com, together with a presentation summarizing the key features of the results.

I would now like to hand you over to the host of today's conference, Mr. Marc Beuls, President and CEO, and Mr. David Sach, CFO. Please go ahead, gentlemen.

Thank you, Operator, and welcome to everyone who has joined us today. For this call, both David and I will be using slides to run you through the results. It will be helpful to have those slides in front of you, so you can find them on our homepage at www.millicom.com. We will both be happy to answer any questions you have at the end, but first I would like to give you an overview of the results and run you through the performance of each cluster.

So turning to the headline numbers on slide number two of the pack, the second quarter of 2007 was once again a record-setting one, surpassing the first quarter of 2007 in terms of both revenue and EBITDA, which grew by 80% and 65%, respectively, year-on-year. We have 1.5 million new subscribers in Q2, giving the total of just under 80 million subscribers at the end of the month of June.

The positive impact of our second billing offer continues. Both in Central and South America, we added more new subscribers in the second quarter than in the first one. Our AAA strategy has allowed us to continue growing market share in Central America, and also in Paraguay and Colombia we saw a similar trend. The margins continue to be strong in Latin America.

As far as Asia is concerned, the increase growth we saw in the first quarter in Asia accelerated in Q2, reflecting the introduction of the second billing in Cambodia and the Tigo launch in Sri Lanka and Laos. Margins increased in Asia from the first quarter of this year, as well as from the second quarter from last year.

Africa had good revenue growth in the second quarter, but the operating margin fell to 31% as a result of a combination of a number of one-off adjustments in the quarter and the cost of improving the affordability of our services that will pay off in the months to come. I expect that the margin will improve from the current level onwards.

Millicom's CapEx for the quarter was $208 million, which represents a 69% increase over a year ago and a 14% increase over the last quarter. And, as you can see, CapEx is increasing throughout the year, as has been the historical trend and we maintain our target of investing over $800 million in the year 2007.

The move to per-second billing in Central America has been an additional driver of growth, and we've been investing in extra network capacity there to accommodate the increase in minutes of use. In Colombia, Africa and Asia, CapEx has been focused on the aggressive expansion of our networks, and as we work to improve the availability of our services, following the launches of Tigo. We look forward to seeing the real benefits of this in Africa and Asia in the coming quarters.

Let's turn to subscribers. As seen on slide number three, Millicom added 1.5 million total subscribers in the second quarter, bringing the total to just under 18 million at the end of June, representing an 84% increase over the second quarter of 2006. In Q2, subscribers in Central America also grew by 84% year-on-year. In South America, subscribers grew by 49%, excluding the Colombian acquisition, and by 183%, including Colombia, where Millicom added over 92,000 subscribers in the quarter.

Africa reported year-on-year subscriber growth of 47% in Q2, and there's every indication that we will be able to capitalize on the growth potential in our popular African markets in the same way that we have across Latin America, as penetration rates rise. The quarterly year-on-year increase for Asia was 44%.

Attributable cellular subscribers increased by 87% from the second quarter of 2006 to 15.3 million at the end of June 2007. It's worth reiterating at this point that we feel confident we can maintain the current run rate of approximately 1.5 million net additions each quarter, despite an expected higher-than-normal churn of customers due to the phasing out of our legacy TDMA and CDMA networks in Latin America in 2007 and early 2008, on which nearly 1 million customers still exist.

We expect that the higher-ARPU customers on these networks will be able to afford a GSM handset and will migrate to our GSM networks. The low-ARPU customers might not be able to afford the handset, however, and we are therefore likely to lose many of them, but their loss from the network will have little impact on revenues because of their low ARPUs.

Phasing out these old TDMA and CDMA networks is important, as it will release spectrum in the 850 and 1900 bands for 3G services, which we are planning to launch in our existing spectrum bands, based on the current licenses in Latin America during 2008 and 2009.

Turning to slide number four, it shows that for Q2 we recorded our highest-ever quarterly revenue of $613 million, an increase of 80% from the second quarter of 2006. In Central America, where penetration is higher and networks are more developed, revenue grew at 49% and Q2 is very strong.

South America, excluding Colombia, produced revenue growth of 58%, which once again represented the strongest year-over-year organic growth for the quarter by any cluster, and reflects the success of per-second billing and the leadership in value-added services by this cluster. Revenues for Africa increased by 46% and revenues for Asia by 35%.

Referring to slide number five, EBITDA for the three months ended June 30, 2007, was $263 million, a 65% increase from the second quarter of 2006, and showing an increase of $15 million from Q1 2007.

The EBITDA margin was 43%, which is very encouraging, particularly in a period of aggressive growth, such as we are currently experiencing. So let's turn to the cluster results, starting with Central America, which today accounts for 44% of group revenue and 54% of group EBITDA.

On slide six, you can see that total subscribers reached 6.7 million, up 84% year-on-year and 13% over Q1. Over 780,000 new subscribers were added in the quarter, attracted to Tigo [novas] and by the per-second billing, which was rolled out across the region in February. This strong subscriber growth soon after the launch of per-second billion has resulted in slightly lower ARPU from deeper penetration.

Second quarter revenue for Central America grew by 49% from Q2 2006 and by 8% from the previous quarter to $271 million, indicating that the price elasticity which we expected to see following the launch of per-second billing, based on our experience in Paraguay, is occurring. EBITDA for Central America increased by 52% to $143 million for the second quarter of 2007, and the EBITDA margin was strong at 53%.

Please turn to slide number seven to look at the results for Central America in more detail. Subscribers grew by 183% to 4.9 million, including Colombia, which accounted for 2.3 million subscribers in the second quarter. In total, looking at the quarterly year-on-year basis, revenues grew by 265% to $188 million, and EBITDA grew by 195% to $66 million.

However, looking at the underlying growth, excluding Colombia, the numbers were still strong, with revenues up by 58%, EBITDA up by 74% and subscribers up by 49%. The EBITDA margin for Q2 was 35%, affected by our efforts to grow our business in Colombia aggressively.

Particularly, we are working hard to grow our market share quickly by improving the accessibility of Tigo. An independent report by A.C. Nielsen in May showed that our distribution network in terms of points of sale inventory was second only to the market leader, and we are quickly closing that gap.

The EBITDA margin in Colombia continues to exceed expectations and in Q2 it increased to 25% from 21% last quarter. Once we have the three A's fully in place, we expect to see accelerated growth in the third quarter of this year and beyond. In the longer term, we expect to see Colombia operate at the average Millicom EBITDA margin.

Turning to slide number eight, in our Africa cluster, there was a year-on-year increase in quarterly revenue of 46% to $106 million, and despite the heavy startup costs of building out the new and extended networks in Africa, particularly in DRC and Chad, where there is a lack of transportation and power infrastructure, we still reported a quarterly year-on-year increase in EBITDA of 15% to $33 million, resulting in an EBITDA margin of 31%.

Total subscribers increased by 47% to 4 million at the end of Q2 2007, following the addition of just under 145,000 new subscribers in the quarter. The results for Africa for the second quarter were affected by a number of factors, which I will briefly elaborate on. First, the promotion in Ghana in Q1 to lower the price of SIM cards attracted many new subscribers, which caused subscribers to call in Q2, as a portion of these new subscribers were not viable long-term customers.

Secondly, aggressive price reductions in Ghana and Tanzania temporarily impacted revenues and EBITDA for this quarter, but it should benefit the third and the fourth quarter. Thirdly, in Senegal, new legislation was introduced in the quarter that requires us to register our customers, and this has significantly impacted new subscriber intake in May and June and will continue to dampen growth in the third quarter.

We expect that the subscriber growth in Senegal will accelerate at the beginning of the fourth quarter. There were also one-off costs of roughly $3 million in Senegal that impacted EBITDA. In Tanzania, we saw the benefits of actions taken by new management in Q2, and we expect this operation to improve its performance in the future.

The beneficial effect of the launch of Tigo is particularly evident in Chad, which has been Millicom's most successful launch ever, and in DRC, where we are pleased with the progress we are making in rolling out the network. Slide number nine shows that revenue for Asia was $48 million for the second quarter, up 35% from Q2 2006.

And EBITDA was $20 million, up 44%, producing an EBITDA margin of 42%. Subscribers in the Asian cluster increased by 44% from Q2 2006 to 2.5 million at the end of Q2 2007. The stronger results for Africa reflect substantial investments that have been made in Asia since 2006, prior to the launch of per-second billing in Cambodia and the launches of Tigo in Sri Lanka and Laos last quarter.

Now, I would like to hand over to David to talk you through the financials.

Thank you, Marc. Please turn to slide ten, where you will see the key financial ratios for the first half of 2007 compared to the first half last year. The slide shows that, compared to last year, sales and marketing costs as a percentage of revenues have increased slightly, but this is what you would expect to see, following a period of Tigo launches in several markets.

The G&A costs as a percentage of revenues are also up, but this is mainly due to the impact of Colombia, and also due to the startup costs in DRC relating to the rollout of the network. Encouragingly, cost of sales have again fallen slightly as a percentage of revenues, demonstrating the economies of scale from growing our businesses and our continued focus on tight cost control.

In addition, expanding our value-added services has helped sustain ARPUs, which in turn had a positive impact on gross margin. Slide 11 shows the recent quarterly EBITDA trend. Following the acquisition of Colombia, EBITDA margins have remained fairly consistent, as the economies of scale from growing our operations have been offset by additional spending on infrastructure and marketing costs.

We are confident that we will be able to trend back towards our pre-Colombian acquisition margin levels over time, as there is room to increase margins in Africa, Asia and Colombia. On slide 12, you will see that there are several items reflecting the profit and loss below EBITDA. And in the next five slides, I will deal with CapEx and depreciation, debt and interest expense and taxes.

The other item worth mentioning is the impact of the sale of Paktel and the disposal of other discontinued operations, which have added $294 million to profits compared to the prior year. Please turn to slide 13. On this slide, we have presented the quarterly CapEx, which shows that CapEx has risen steadily throughout most of the last six quarters. This trend is expected to continue, although not as steeply as in 2006.

We have reported CapEx for Q2 2007 of $208 million, bringing the amount for the first half of the year to $391 million. Given the upward quarterly trend, we reiterate our previously stated forecast of over $800 million of CapEx for the full year. Going forward, we expect the CapEx to sales ratio to begin trending downwards.

As Marc has already mentioned, the focus of our CapEx spending in Latin America has been on capacity expansion to accommodate the increase in minutes of use, resulting from the introduction of per-second billing. In Colombia and Africa, and to a lesser extent in Asia, CapEx spending so far this year has been focused on expanding our networks.

As many of you know, the payback periods from capacity expansion are relatively quick, since equipment prices have been falling for many years, whereas the payback periods from coverage expansion are longer, because the prices of the [silverworks] have been rising. We believe that the current blend of coverage and capacity expansion is similar to previous periods, and will enable us to maintain our historical payback periods and rates of return on our investments.

Slide 14 shows the quarterly breakdown of deprecation, which indicates that depreciation has increased steadily due to the rising levels of investment that we are making in our businesses. This trend will continue for many years, as depreciation will converge towards the average CapEx spending of the past six to seven years.

Currently, CapEx is running at a rate of over $200 million a quarter. Despite this trend in absolute values, depreciation as a percentage of revenue has remained fairly constant, and we expect it to stay flat, even as we start to spend aggressively on converting our networks to 3G technologies.

On slide 15, which breaks down our quarterly debt, you can see the debt has settled at around $1.5 billion, following the Colombia acquisition, and has risen slightly since then, and now totals $1.6 billion. Due to our underlevered position, with net debt to EBITDA of roughly 0.5 to one, we will increase debt to finance our CapEx spending in the operations. This is likely to increase our cash balance even further.

As mentioned on the previous quarterly call, we are reviewing our capital structure to determine the most appropriate level of debt, and to ensure that we utilize our cash in the best interest of shareholders. We are considering further minority buyouts, looking at potential acquisitions and greenfield site startups, seriously considering the merits of a share buyback, and will look to repay the corporate 10% notes in December 2008, when they become redeemable.

Please turn to slide 16, to see the corresponding quarterly interest expense. The simple effective interest rate has stayed relatively stable throughout the last six quarters, at roughly 10%. We are looking at ways to reduce the average gross interest rate, but more importantly believe we can reduce our post-tax interest costs by borrowing more at a local level, where the interest is usually tax deductible, and less at corporate, where we have been unable to deduct the interest.

For this reason, we are very likely to repay the corporate 10% notes when they become redeemable next year. Slide 17 shows that the tax rate has risen year-on-year despite our initiatives to better manage our overall tax position. We have been able to reduce the non-deductible net corporate expenses slightly by increasing the management and brand fees to offset the higher corporate costs.

However, as we anticipated, the operations tax rate has risen due to the operating losses incurred in our business in Colombia and DRC, which provide no tax benefits. The operations tax rate is likely to stay at the 28% level, or might even rise further this year, as these companies represent a larger proportion of profit before tax for the total group. Over time, this tax rate will likely fall, as these businesses become profitable.

From 2009, we may be able to get the overall group tax rate to 30% or below, but this will depend on the geographical mix of profits, the profitability levels we can achieve in Colombia and DRC, whether we repay the corporate 10% notes and our ability to transfer significant income to corporate to fully offset the remaining non-deductible costs.

On slide 18, we show the strong operating cash flows, which are more than financing the increasing level of investments within the group. Helped by the sale of Paktel, we ended the period with a very strong cash position of $1.1 billion. Our net debt position is $549 million, representing a ratio of 0.5 to one, against extrapolated full-year EBITDA.

With such low levels of debt, Millicom is well placed to continue its significant investment program, and as mentioned previously, is considering the best use of its cash to maximize shareholder value. Slide 19 illustrates the continuing strong investment across our businesses. CapEx was $391 million in the first half of 2007, compared to $203 million for the first half of 2006.

As mentioned previously, we expect to see absolute CapEx figures to trend upwards throughout the year, and our current expectations are that CapEx will exceed $800 million for the full year. Thank you.

Now, I will hand you back to Marc.

Thanks, David. To summarize the past quarter and looking forward, I can say that Q2 was a strong quarter again, and that we expect the year 2007 to be another record year. I feel confident that we can continue adding 1.5 million subscribers a quarter in 2007, based on the planned CapEx spending. In Latin America, we continue to improve our market position, despite the increased competition.

The introduction of per-second billing has improved the affordability of our services, leading to strong subscriber intake. The focus going forward will be on value-added services, one of the reasons why we are planning gradually moving to 3G within our current frequencies and licenses in the years 2008 and 2009.

We hope to see you all in Colombia for our IR days, starting October 29th, as we are building an interesting growth story on the back of much improved accessibility of Tigo prepaid services. If you would be interested in attending, please contact Shared Value.

In Africa, we are facing challenges such as the lack of basic infrastructure that at times will delay our aggressive expansion plans and temporarily impact our profitability, given the higher OpEx. I continue to see great opportunities in Africa where the telecommunications infrastructure is still poor and the mobile penetration low. Consistent implementation of our AAA strategy will allow us to take advantage of those opportunities.

In Asia, we have improved our service offering by introducing per-second billing in Cambodia and launching Tigo in Sri Lanka and Laos. The growth rates in those regions have started going up, and we expect to see Asia performing in line with the two other regions.

That concludes my comments, and we will now be happy to take your questions. So, Operator, may I have the first question, please?

(OPERATOR INSTRUCTIONS)

We will now take our first question, from Bill Miller, from Hartwell Investment. Please go ahead, sir.

Marc, just a couple of questions on the statement you started out with in your press release, to -- increased competition. Could you characterize that increased competition, because you've always been in competitive markets? That's the first question.And, secondly, if it is increased in some way, shape or form which you're going to tell us about, would you tell us what your reaction to that increased competition is?

The immediate impression I got is that it was going to impact margins. From what you say on the call, it's not going to impact margins. Can you tell us what impact it will have? It doesn't seem to be on the growth rate, but as you look forward, can you see, as you've said here, increased margins?

The second question is about your CapEx. Historically, you've been able to visualize a 50% two-year payback on your CapEx. Is that still the case, or are we running into something there which is also reflected in the competitive environment? Thanks a lot.

We talk about increased competition. It's not that there are a number of new operators coming into the markets. I think everybody knows the markets where we expect new operators to come in, places like Honduras, Senegal, where there are already two operators at this point in time.

When we talk about increased competition, it's that by us having introduced our AAA business model, we have seen that the competition has started reacting. The impact, as you've correctly said so far, has been negligible. As a matter of fact, for Latin America region, especially Central America and Colombia region, I visited not less than a month ago, we continued to increase our market share. Our margins continue to be very strong in Colombia, even going up.

So we don't see an impact on the financials at this point in time. We see that people, as we've always said, copy our model, but they don't always copy with precision. So no impact on margin, no impact on growth at this point in time, and we don't expect that that will have an impact going forward.

In terms of the CapEx, given that our margins are not affected by this event, we don't see any change in terms of the paybacks we are seeing for our CapEx, which are like two years for existing -- investments in existing businesses, and could between three to five years for investments in new businesses, so none of those things are being impacted as we speak.

I guess, one further question. On the ability to raise money in the local markets, is that still available to you as we speak, and do you expect the depth and breadth of those markets to continue to be as strong gas they are?

Local debt continues to be available in the markets, in Latin America, but also now more in Africa and Asia. So we will continue using those pockets of money at the financial markets, as the debt markets are improving in those markets, we will be able to borrow more locally. So that is not a constraining factor for us. As a matter of fact, every year, we're able to tap more into the financial markets, more into the debt markets, than we were able to do the year before.

Well, that sounds terrific, I'll let someone else do the [question]. Thanks very much, Marc. Good quarter. Bye.

Our next question will come from [Drake Johnson] with Davenport. Please go ahead, sir.

Hey, guys, how are you doing?

Hi, there.

Good.

A question for you. You had made a comment in Colombia, where you had seen your margin improvement 21% in the first quarter, 25% in the second quarter, that you expected Colombia to get to the margin you have for the firm. Do you have a target for that?

Are you looking out 18 months, or do you expect continued improvement in Colombia margins this year? Because after the first quarter, you basically said you thought the Colombian margins would remain where they were at the remainder of the year, but it sounds like you are making better progress there.

What we said about Colombia is that our focus is our top-line growth, and that's why we have been focusing a lot on improving the accessibility or the distribution of the product and the services, and you've seen the results of the A.C. Nielsen report. So clearly, the team locally has done a great job in terms of improving the visibility and the distribution -- the visibility of Tigo and the distribution. And again, I saw it with my own eyes when I was there a little over a month ago.

So that is working out, and we expect that the third and fourth quarter of this year we will start seeing good revenue growth. The margin improvement, we're extremely happy with it. It's probably a little bit higher than what we anticipated. So what, great news.

And I think it has -- it is the result of some acceleration of the revenues we start seeing towards the end of the second quarter, and that's why we feel confident that the third and fourth quarter is going to show a lot more revenue growth than we have been showing in the first two quarters of this year.

And then in Africa, you have highlighted some of the challenges there and you have mentioned that in the second quarter you took a $3 million charge. If we look at the operating pass through margin on Africa, excluding that $3 million charge, I believe it was roughly 34%. Do you actually expect that margin to improve in Africa over the remainder of the year?

I see no reasons why the margin would not go up. There were a number of factors, some one-offs in the quarter, some kinds of price reductions we did, as we did in Central America in the beginning of the year in order to improve the affordability of the service and given that some of the price levels in Tanzania and other markets were still fairly high, so we felt that we needed to bring them down in order to fuel growth.

And as far as Tanzania is concerned, and a country I visited three weeks ago, we start seeing the first positive impact of those price reductions and improved distribution visibility. We expect to see similar things in Ghana and the other African markets in the course of this year. So the margin I think can only go up from the level it was in the second quarter.

Marc, just a last question. The tax rate in the second quarter, your blended tax rate, looked like it was about just under 30%. I see it was almost 37% in the first quarter. So you were sort of comparing first-half rate of '07 versus first half '06.

I mean, is there any potential that your third and forth quarter tax rate this year might remain in the 30% level, or are there factors -- you had mentioned -- depending on the profitability in some countries, it might affect your tax rate.

Yes, I mean, the thing affecting the tax rate most of all these days is Colombia and DRC. Both obviously, they have losses on a pretax basis and we do pay taxes, so there's tax there without any profits, so that obviously drives up the overall tax rate. I mentioned on my part of the call that that could go up a little bit this year, depending upon the CapEx spending depreciation in those countries, but I don't expect that to be significant.

So I would think the tax rate in the low 30s would hold, maybe go up slightly depending on the losses in Colombia and DRC. And also the other impact obviously is corporate, and our ability to move income into corporate, which we've been quite successful at. And you can see on the tax slide that I showed you, we've been able to offset a lot of that pretax losses at the corporate level with the additional income. So all things considered, I would expect the tax rate to stay in the low 30s.

Low 30s, and then going -- in '08, it might stay there and potentially actually drop down when you -- when we move the debt from a corporate level and move it to the subsidiary level or pay it off by the fourth quarter of '08.

Correct, and then obviously, as Colombia and DRC start generating profits, that will affect the overall tax rates, as well. So I would expect that that operations tax rate of 28% will come down slowly over time and then also the corporate effect that you mentioned. Correct.

All right, thank you very much.

You're welcome.

Thank you.

We will now move to Anders Wennberg from RAM for his question. Please go ahead.

Hello, Anders Wennberg from RAM. First back to the competition in Central America, we've seen ARPUmoving from about 25 down to just about 20 in two or three quarters. You at the same time introduced per-second billing, you explained before, but should we expect ARPU to stabilize at this level, or do you still expect a rebound, or is the most likely scenario that competition will keep increasing, ARPU keep dropping?

Let me first say that the drop in ARPU doesn't have anything to do with competition. The drop in ARPU has to do with the price reduction we introduced on February 7th of this year, okay, where we introduced a 25% price reduction on purpose, because we needed to constantly improve our affordability in order to sustain the growth. Okay?

And if you look at the subscriber growth that we generated in the second quarter, which as I said on the call was higher. The net subscribers in the second quarter was higher than in the first quarter. By the way, it was the same in South America. Okay? So as a result of the actions we took in the beginning of this year, we see an acceleration of the growth.

So the ARPUs have nothing to do with competition. This has to do with a voluntary price reduction, which we think is necessary to continue fueling the growth in the medium to long term. And that's also the reason why we're doing this in Africa. That's why we're ready to take a hit in the quarter in Tanzania and Ghana, where we brought down prices, in order to make sure that we continue fueling the growth.

So Claro has not followed in terms of per-second billing and --

I didn't say that. We are, as I said earlier, we are being copied by our competitors. That was the case in the past and, as we said in the release, we are being copied faster nowadays. But judging from the market share numbers I saw in Central America and South America, judging from the margin we have, judging from the subscriber intake, it is not impacting our performance.

One could even say on the contrary, we are increasing our subscriber intake, our margins are 53% in Central America, a very strong 53% in Paraguay, increasing margins in Colombia. So I don't think that it's impacting our financial performance at this point in time.

Second question, if I may. I came in a little late. If you have already commented on it, you can skip it, but what is the process in Peru regarding the new license, and also new licenses in Paraguay and other countries, potentially Costa Rica, et cetera?

We're not involved in Peru, and I don't think the processes in Panama and Costa Rica have officially started at this point in time, but these are companies, as they are close to existing markets where we will keep an eye on and probably get involved in the process, if and when it starts.

Okay, thanks.

Thanks, Anders.

Our next question is coming from Bengt Molleryd with Standard & Poor's. Please go ahead.

Thank you. Bengt Molleryd from Standard & Poor Equity Research in Stockholm. I just wondered, when it comes to Africa, has anything changed on your view in the market, the dynamics in the market now with the subscriber numbers ticking downwards and people leaving? Are you surprised in any way to the volatility among the customer bases in the different markets?

I don't think any of the fundamentals in Africa have changed over the quarter. I mean, fundamentals definitely don't change that quickly. What we see is a (inaudible) opportunity, but it's an opportunity where there are a number of challenges and the challenges are seen in the form of the delays we see rolling out networks, the difficulties getting equipment on place, stuff like that.

But in terms of what it takes in order to build a successful business, it is the same as we did in Central America and Asia. Also there, at times, we run into bottlenecks, but we always managed to resolve those, and once we were past those bottlenecks [know], it was business as usual.

So I don't think that anything has changed in Africa. We have now the networks in place. We will continue building networks, and then with our much-improved offering, and so it is key that we continue to improve our affordability, that we will be able to continue growing those markets.

I don't think we would have been able to continue growing the African markets at the price levels we have prior to the price reductions in Ghana and Tanzania, for instance.

If I may, are there any severe or more difficult factors in building a network in Africa there that delay deployment or emission or electricity or spare parts, or civil engineering, other factors that are making it more difficult for rollout the network in --?

The civil engineering is getting concrete on the site when there are no roads can be challenging. Getting power is challenging. At times, it's impossible, so you have to run on generators. So you run on fuel. It has a higher operating cost. Those are factors that we see more in Africa than we've seen anywhere else in our operations. Having said that, the economies where we are currently operating are growing extremely fast.

And what you have in fast-growing economies is that sometimes the infrastructure cannot cope with the growth, so you get bottlenecks, and power is typically one of those bottlenecks. So those are the things we have learned to live with but kind of delayed at times the rollout of networks. And, as we said before, it impacts profitability or can temporarily impact the profitability of our business. And that clearly is what happened in the second quarter this year.

But you said that you expect margins to come up going forward during this year. Are the factors improving rapidly here, or what is behind that -- the more improved margin in the latter part of this year?

The improved margin in Africa is going to be possible as a result of, first, a number of the issues were one-off, so we don't expect them to come back. They're non-recurring. Secondly, we expect that the reduction of tariffs we introduced in Tanzania and Ghana have improved the affordability and will allow people to buy more minutes, and will allow more customers to have access to more of our services.

So we think that the payback is going to be relatively short -- payback of those price reductions is going to be relatively short, as have shown in other markets where we did similar things, like in Latin America, South America first, and Central America later.

Thank you very much.

Thank you.

We will now take a question from Kevin Roe, with Roe Equity Research. Please go ahead, sir.

Thanks. Hi, Marc and David.

Hi.

Hi, Kevin.

Hi. On Africa, if we adjust for the one-time costs in the quarter in Africa, do you still expect to see margins to be up sequentially?

Yes, we do. Yes.

Okay, and can you give us some more detail on what those one-time costs of $3 million in Senegal, what that comprised of?

The one time was the write-off we had on [VET] receivables, which we thought we could recuperate, but at the end of the day it was not possible, and that's why we took a write-off.

Okay, and in Ghana you mentioned the SIM card issue and that affected Q2 subs and the subs fell sequentially. Has the subscriber base stabilized, or are you now seeing net subscriber growth in Ghana? Is that adjustment behind you?

I think that adjustment is behind us. What we learned from that, and this action was run in the framework of the 50th anniversary of the independence of Ghana, and so we wanted to be friendly to the population there by giving free SIM cards away, but it shows that you want to get subscribers on the network that show some stickiness and show loyalty.

And that generates revenues going forward, and that is what we are focusing on today. So we expect that we're no longer going to have those fluctuations going forward, but it also shows that our subscriber base in Ghana now is a solid, clean subscriber base, with revenue-generating customers.

And a couple more questions. On Colombia, is the introduction of per-second billing in the cards for this year, or is that a later event?

I'm playing poker right now, so I won't show you my cards, Kevin.

Okay, and lastly, your $1 billion-plus cash position, you mentioned in the prepared remarks, share buyback, greenfield, acquisitions, minority buyouts. Can you give a sense of -- can you prioritize your preferences there, and when do you think the board will come back to shareholders with the final decision on a share buyback, if any?

I can't answer that question, so you'll have to ask that question to the board. In terms of our top priority is buyout of minority shareholders, as you know, followed by getting started in some markets that are adjacent to the markets where we currently are active in Latin America and in Africa. So that's where my priorities are.

And are you making any good progress on minority buyout discussions?

The discussions are ongoing, but they have been ongoing for a number of years, at times, so it's very hard to predict whether this is going to happen, and if so, when it's going to happen.

Very good, thank you.

Thank you, Kevin.

Our next question comes from Andreas Joelsson with SEB. Please go ahead, sir.

Good afternoon. It's Andreas Joelsson from Enskilda in Stockholm. I have a question on Congo. It's a quite significant drop in net subscriber intake from Q1 to Q2, and I was wondering what the reason for this is, and do you expect to come back to the levels seen in Q1 already in the second half of '07?

Well, what we saw in the second quarter of this year is that, following the relaunch of our business in January of this year is that the cards we had distributed to some resellers, resellers were not selling in the market but they were selling them to one of our competitors, who would change them for one of their SIM cards. So we have SIMs -- activated SIMs, on the network that were not reloaded. So the minutes on the SIM were use and then that was it, there was no reload.

And that's what we needed to correct in the second quarter, and that's why the growth in Congo was not as strong in the second quarter as it was in the first quarter. So this is out of the numbers. We've changed the way we deal with our resellers, in such a way that they get most of the money not upon the sale of the SIM, but on the first reload of the account, in order to avoid that kind of practices.

Okay, thank you.

Thank you.

Our next question will come from Peter Nielsen with Cheuvreux. Please go ahead, sir.

Thank you. Peter Nielsen from Cheuvreux. A couple of questions please. Firstly, of the TDMA, CDMA customers which you expect to churn in Central and Southern America, how many of these have been churned in this quarter, if I may ask? And secondly, just out of curiosity, the price adoptions in Africa you talk about, again, in Tanzania, just -- what's the size of these reductions, if you would tell us roughly? And when about in the quarter they were made?

And thirdly, really into the CapEx guidance of sort of above $800 million, as you say, we would expect CapEx to accelerate in the second half, and we are almost at the $400 million, so what' the likelihood that it will be sort of significantly above $800 million? Will you give us sort of a cap for where you see CapEx hitting this year, given no new market interests? Thank you.

In terms of he TDMA, I don't have a number for you there. I mean, this is an ongoing churn, so as you can see yourself, it doesn't really impact the revenue. I don't have a number for you there.

In terms of the price reductions in Africa, yes, they were substantial price reductions, I would say between 10% to 20% at times, and they were made in the quarter -- the second quarter. I can't remember the exact date those price reductions were made, but they did impact quite a big part of the quarter.

In terms of CapEx, yes, historically, we always increase our CapEx forecast. I think in the meantime, I think we've gotten a little bit better in terms of forecasting our CapEx, so I don't expect a big increase in our CapEx forecast. I mean, there can always be some small adjustments in the fourth quarter. Historically, we used to have a big increase in the fourth quarter.

I think nowadays we spread our CapEx more across the whole year. So don't expect a massive increase in the CapEx towards the end of the year. As a matter fact, for the revenues we want to make in the fourth quarter of this year, that CapEx is already in progress, so that's already being shipped or being installed on the ground.

Okay, thank you, Marc.

And our next question will come from Stephen Mead with Anchor Capital Advisors. Please go ahead.

Yes, Marc, can you just go into a little bit more detail as far as the revenue benefit from moving customers from the CDMA to the GSM? And then also, what impact on CapEx this transition to the GSM in terms of the spending this year and next year? And then, as you look at sort of the profile of the customers that are on the sort of upgraded GSM system, what is a typical customer -- kind of look like, and how big is that market that you're going after?

Okay. Well, in terms of the move from TDMA to GSM, historically we've seen that ARPUs are higher for the simple reason that a number of value-added services that are available on GSM networks are not necessarily available on TDMA networks. So from that point of view, it's a beneficial move for us, as well as for the customer.

So in terms of the CapEx spending, we are amortizing our TDMA and by the end of the year, most of that will be out -- not all will be out of our books. And you know that the next generation of technology always comes at a lower price than the previous ones. And by the way, we expect that that third-generation, now five years after it was introduced, is going to come at a lower price per line than GSM equipment, and definitely at a much lower price than TDMA and other technologies we were using previously.

So technology changes, especially as we can do in Latin America, within our existing spectrum, so we don't need to buy any new spectrum. We don't need to get any new license. We just reuse that same spectrum, but using a different technology, now it only comes at a lower cost.

In terms of the profile of the people that might not kind of make the move, I mean, it's difficult to say because this is prepaid business, so we don't necessarily know those people. But these are people with very, very low purchasing power, people who are probably using very old phones, phones that might have been given to them, and they're just going to afford basically the reloads of the minutes, but might not be able to afford the phones.

So maybe we're a little bit too negative here. Maybe by the time they need to find the GSM phone -- there are going to be those old GSM phones at very low prices available and maybe some people will give them those phones for free.

So maybe they will be able to make the move, but we know from previous technology changes that there is always kind of a tail of subscribers that don't make the move or maybe people who are holding two phones and just disconnecting one and continue with the other. Who knows.

What's the difference, though, in ARPUs, though?

We're talking about ARPUs of people at ARPUs below $1, so these are very, very low ARPUs. And are those people going to be using value-added services if they were to get a GSM phone? Probably not. So when I talk about higher ARPUs, it's more like saying they're higher-ARPU users. When they get a GSM phone, they typically spend more on their phone than they used to do before.

And then in Central America, in terms of share gain versus growth of the market, where do things stand in terms of the growth that you can expect in Central America?

We continue to see very strong growth, and we've been gaining market share in all three markets over the last six months. And we're now the clear market leader in Guatemala with a market share at or maybe even slightly over 40%, so we've become the market leader. In El Salvador, with five operators, I think we increased our market share by 1%, and also on Honduras, we increased our market share by 1% or 2% over the last six months.

So this AAA approach, this per-second billing we introduced, has a very beneficial impact on our market position and it doesn't impact our profitability, as you can see from our high EBITDA margins in that part of the world.

But are those markets pretty much at saturation?

No, no. I still don't know that word, and we see mobile penetrations in Latin America today reaching 80%, 90%. Places like Chile, Argentina, maybe even Mexico, they're all moving in that direction. I think we will be able to move very closely to those levels also going forward.

All right, thanks.

Thanks, Stephen.

And our next question will come from Alexander Vassiouk with Morgan Stanley. Please go ahead.

Yes, hi, a question on elasticity in Central America. Could you just maybe comment on how the volume of traffic continued to develop during the second quarter? I remember in the first quarter you mentioned that traffic was up 25% towards the end of the quarter in response to moving to per-second billing.

Can you give an indication of the growth rates that you were seeing during the second quarter? Any data points would be very useful, just to understand the underlying dynamics.

I don't have any hard data for you there, but we know in this industry the higher market share you have the more on-network traffic you have, and that is beneficial for your margin. So, as I said a few minutes ago, we have continued increasing our market share in those markets, so probably there is even a better balance when it comes to the traffic in that region, but I don't have any hard data for you right now.

Okay, that's fine. Thanks very much for that.

(OPERATOR INSTRUCTIONS)

We will now take a question from David Kestenbaum, with Morgan Joseph. Please go ahead.

Marc, can you talk about why ARPU went down in Central America versus went up in South America during the quarter?

I think Central America went down because of the very high subscriber intake we saw. As I said before, our subscriber intake in the second quarter was again higher than in the first quarter, so in that, it has a dilutive impact on ARPUs.

I think in South America we see a combination of very strong offerings of value-added services in places like Paraguay. They push ARPU -- I think ROI is also benefiting from a very strong local currency, which will also positively impact the ARPUs, of course. And then of course, Colombia. We see very strong performance there, with revenues that really start going up now. We cleaned up our subscriber base there. So the subscribers we have there are revenue-generating subscribers.

So I think that explains the difference between South and Central America. But you've seen also yourself that the subscriber increase in Central America is a lot higher than it is in South America. And by definition then, your ARPUs are lower.

Okay, can you talk about how many points of sale you have in Colombia now and where you think you need to be?

The number of points of sale in Colombia, I think you have to come down in October, David, to count them yourself. But we have increased that number. Just I don't have the number on top of my head, but we've increased the number substantially and in terms of prepaid points of sale, we're now number two in the market according to that A.C. Nielsen report.

And if I remember, we are present in 11% of all points of sale in the country, compared to 12% for the market leader in the market. So for any kind of point in sale in Colombia, we are present with inventory in 11% of those.

So you're not far behind then. And then having the minority partners in Central America, has that affected your ability at all to compete and react to the competition?

Not at all. Not at all. We have full understanding with our local partners as to what it takes to keep ahead of competition. That's the reason why we decided jointly to introduce Tigo, to introduce jointly per-second billing, e-pin and all of those things. There's no conflict whatsoever. There's no difference of opinion whatsoever as to what it takes in order to grow that business.

Okay, thanks.

Thank you.

Our next question will be follow-up question from Andreas Joelsson with SEB. Please go ahead.

Actually, my question was just recently said by my colleague here, so sorry for that.

No problem.

And now we'll move to Bill Miller with Hartwell Investment. Please go ahead, sir.

Marc, hi, sorry to bother you yet again, but I'm still somewhat confused by your fourth line, when you talk about increased competition affecting you. I am therefore going to ask the question about the transferability of the AAA marketing strategy, a.k.a. Tigo, to Africa and these other markets. It sounds like it's been very successful, but could you shed a little light on that, perhaps, and tell us that it is or that it isn't transferable very easily?

Well, it is transferable, and that's what we showed to those who joined us on the trip to Ghana last year in November. And then if you would go today to, let's say, Dar es Salaam, Tanzania, you would see that Tigo is as visible and as present in the street as it is in Tegucigalpa in Honduras.

So, yes, it is transferable, and that's also the reason why we've been able to grow the revenues in Africa just under 50% in the second quarter. I'm sure that without that AAA strategy, we would not have been able to do so.

So I think the transferability is there, at least in house. I think how easy it is to copy it by a competitor, I think that is probably more difficult, not impossible, but more difficult, but in house it is doable. And to give you some examples, the guy who is running Tanzania today is a guy coming out of Paraguay. He has run this in Paraguay, the Tigo concept is now implementing it in Tanzania.

One of our cluster managers in Africa comes out of Paraguay, so he had done this in Latin America and is doing it now in Africa. And that's the strength of Millicom, where we are able to find people who have done it, and then take that expertise in the next market and implement it there in exactly the same way.

Well, it sounds terrific and I guess I'm just still confused by that opening paragraph. What percentage of the Congo is built out now?

Well, it's difficult to quantify that, and I think we're now over 300 -- I think, 310, 320 base stations from -- I think just over 200 when we launched in January. So I think we kind of increased our number of base stations by 50% over the last six months. So we're making good progress, but we need to do more. How much of the population that is covering? I don't know exactly.

Well, it sounds terrific. What is your target for that?

Ideally, we would like to get over 400, 450 base stations by the end of the year, and then we will continue building out next year. So this is an ongoing process.

Well, it sounds terrific. Great, thanks.

Okay. Thanks, Bill.

(OPERATOR INSTRUCTIONS)

Our next question will come from Sven Skold with Swedbank. Please go ahead.

Hello, this is Sven from Swedbank. Just a question on margins again and back to Central America. Margins in Q2 in Central America were higher than last year in Q2, but they were lower than in Q1. Is that only because you do add -- or, what's the explanation for that?

I think as you've seen yourself, the subscriber intake has been a lot higher in the second quarter than in the first quarter, and by definition now has an impact because of the subscriber acquisition costs. There has been no change in any of the fundamentals in the market that explains any major shift in the cost of doing business in Central America.

So the gross margin is stable or even up then, because volumes are higher.

Yes, volumes are high, gross margin is stable. But, as I say, the more subscribers you acquire, the higher your cost is, and I think we did about -- is it 60,000 or 70,000 subscribers more in the second quarter than we did in the first quarter.

Sure. If that's the only explanation, that's fine.

That's the only one I know.

Great. Thanks.

Thanks.

(OPERATOR INSTRUCTIONS)

We will now take our next question from Bengt Molleryd with Standard & Poor's. Please go ahead.

Bengt Molleryd from Standard & Poor. Just to follow-up, have you anything new to report regarding Vietnam regarding the license or potential your involvement there?

Nothing new, Baint. The process of privatization is ongoing and the government is in the process of selecting a financial adviser, and we also know there is [something] going on within the government, some ministers being replaced by others. But no change in our position in Vietnam.

Do you think it's possible to see an equitization process later this year, or should we expect that to happen in 2008?

I don't expect that any significant things will happen this year. Probably the choice of the financial adviser, and I think that's where probably it will stop for this year.

Thank you.

Any more questions, Operator?

At this time, we have no questions and I would like to turn the call back over to Mr. Beuls for any remarks.

Okay. Thank you, everybody, for participating in today's call. If you have any other questions, feel free to call David or myself or Andrew Best at Shared Value. And again, thank you for being on the call and have a great day. Goodbye.

That will conclude today's conference call. Thank you for your participation, ladies and gentlemen, and have a nice day.