Ladies and gentlemen, thank you for standing by, and welcome to Axiata Group's third quarter 2019 results briefing. Throughout the presentation, all participants are in a listen only mode. There will be a presentation. There will be a presentation followed by a question and answer session. There are please three housekeeping reminders. Please mute your phones during the presentation and kindly avoid using wireless headsets. Also note that the call duration will be for a maximum of 90 minutes, ending at 5:30 P.M. I would now like to hand the conference over to the Chief Executive, Tan Sri Jamaludin, President and Group CEO. Thank you, sir. Please go ahead.
Hi, this is Jamal. Thank you for joining us for the third quarter 2019 results. Let's go straight to slide four. These are six, seven messages that we want to convey. The first one is, of course, perhaps the most important, is consistent with our focus on profitability and cash. I think we have seen more and more definitive results. As you can see, the year-to-date free cash flow has jumped 33.5% to MYR 2.6 billion, boosted by the operating leverage from higher revenue and cost excellence. Our underlying PATMI, of course, affected by the absence of M1 contribution and higher taxes for Robi, which is quite unfortunate, but still has been quite good. You exclude those items, the year-to-date PATMI increased by 18.6%. Our balance sheet remains very strong, which later on my CFO will elaborate.
The one thing we want to highlight is, and we are very happy with, is the XL continued performance to show that the turnaround is real. As you can see from the performance, the ARPU increased and the revenue increased by 10.6%, and the return to profit with ROIC improvement of 4.5 percentage points. Robi. Robi was profitable, after the acquisition of XL, of course, they have absorbed XL negative EBITDA and profit. We are happy to say that the return to profit with a strong improvement in ROIC at the same time. Celcom delivers higher PATMI of 7.9% and EBITDA of 15.3%. We are also happy to note that Axiata continued with a double-digit growth in all metrics, and that's something that will help Axiata grow. ADA, our digital advertising company, delivered net income profit in third quarter.
We are excited for the year to be positive. As promised, or as we had conveyed earlier, there are three companies within our digital businesses, ADA is one of them. Our expectation is that to deliver profit for the full year, and they are on course to do that. On the headline KPIs, barring unforeseen circumstances, of course, we are likely to achieve our FY 2019 guideline KPI for EBITDA and ROIC. The next slide five, is just to give you a flavor, and when we say we want to improve EBITDA, PATMI and FCF, we achieve in practically all the companies except for MCell. As you know, MCell, the ILD, International Other Businesses, is not performing as good and as expected for most parts, right? Of course, the introduction of some new taxes in Afghanistan affecting consumer demand has affected them.
I would say that it has somewhat affected them. On the relative performance of the group, and all the operators as you can see, in all metrics we perform as the number one and number two. The next slide six, is just to show that we have increased our revenue practically across the board, except for maybe Celcom and MCell. As you can see, correspondingly, the total cost is lower in practically all except MCell. For example, for the group, we increased revenue by 5.5%, but our total cost increased only 1.5%. I will pass over to my CFO, Vivek, who will give you a bit more detailed explanation on the results. Vivek?
Thank you, Tan Sri. A very good afternoon to all of you. I'll just go on to slide number eight. A quick rundown on the financial performance for quarter three 2019. First slide talks of the reported numbers, and the second next one would be on the underlying performance. Reported performance, revenue grew strong at 4.0% and EBITDA at 26.3% on a year-to-date basis. 26.3% was coming out of contributions from the MFR as well as exchange gain in 2019. PATMI overall delivered year-to-date over 1 billion, coming out of a few one-off items like M1 gain that we sold in February this year of around MYR 118 million, divestments, of course, our non-core digital business of MYR 302 million, and the Idea rights which we have given up, which they gave us around MYR 96 million of gain. Overall, the quarter numbers were strong.
If I go to the underlying performance, which is really what matters, which is on pre-MFR basis at constant currency, having a strong growth in EBITDA, a double digit 10.3%, despite a 2.4% drop in revenue. If I exclude device sales, our service revenue excluding devices grew by roughly around 1.1%. This resulted in 2.3% point improvement in EBITDA for us at 37.8%. Costs more or less remained same flat as Tan Sri said on the earlier slide. PATAMI was dragged by a few issues or items of one-off. One is we sold off M1 during the year, which was contributing, I mean, proportionately for three quarters had contributed MYR 19 million last year, which we won't get this year. The taxes which were introduced in the budget of Bangladesh, which is on the revenue increase from 0.75%-2%, has an impact of around MYR 84 million for us.
Then some of the M&A expenses, which is relating to the discussions we have with Telenor, would be a net impact of around MYR 10 million. Those are the items which really impact the underlying PATAMI for the first three quarters. However, as Tan Sri mentioned earlier, if we normalize for some of these one-off items, it will be around 13.6% improvement in profits. The next slide basically gives a waterfall from last year to the current year, and I think some of it I already mentioned. EBITDA contribution is significant for us in first three quarters, followed by an impact which is coming largely on taxes other than the M1 which I talked about. Taxes mainly coming out of Bangladesh, which has a dual impact. One is the corporate tax increase, which I talked about 0.75% to 2%, but also impacts the deferred tax accounting.
In addition to that, there's a tax implication in XL, which was on the carried forward losses coming from the AXIS merger, which part of it expired this year. In addition to that, I think you see the finance costs slightly higher, mainly because we have moved specifically in Indonesia, where we've moved from USD borrowing to the local borrowing, which has a much higher interest rate, but that's to protect us from forex fluctuations. If I go to the one at the bottom, basically gives the waterfall between the reported number and what is the underlying number. The big item there is the M&A gains, which I talked about earlier, coming from the M1 disposal, Idea rights issues, and Idea non-core ventures, which we carved out during the year. If I go to the next slide, our focus has been on improving our cash flows.
I'm happy to say that we've seen a positive impact of around MYR 2.6 billion on our free cash flow. If I look at it on a core semi-cash basis, which the table shows below, our free cash flow margins are 19%, and operating free cash flow margin improved from 3% to 9%. Our CapEx intensity remains very much similar at 24% compared to last year. Mostly coming out of EBITDA improvement, you will be able to see cash flow improvement. Operational excellence are the key focus for us. We are trying to keep efforts on improving the cost, despite the fact there is expansion happening in some of the markets. For example, Indonesia, Robi, Bangladesh, [inaudible], Celcom, we've added new sites, which has had an impact on increasing the network cost.
Through our efforts on the cost optimization program, we've been able to reduce our cost on the network side and the marketing staff, mostly coming out of a few of our companies, XL, Robi, and somewhat in Celcom. Broadly, the costs have remained flat for us despite the expansion in the network. If I go to the next slide on the balance sheet, a very fairly strong balance sheet with the gross debt to EBITDA down to 1.9 on a comparable basis. We did move out to higher books with debt and moving more into local currencies last year because of the volatility in the market. That does have some marginal impact on our cost of borrowing.
We do feel that given the volatility in the currency market as well as interest rates, it's much more prudent for us to have a higher fix and local currency borrowings. Our cash position remains fairly strong at RM5 billion. If I go to the next one, Celcom. I think market has been struggling a bit in terms of the market build overall, but largely coming out of two factors. One is the wholesale revenues which are coming down, as well as the interconnect rates, because interconnect rates coming down is indeed an effect on the industry growth. As far as telecom is concerned, I think overall growth has been in line with how the industry is doing too. Our core revenue has been positive at 1.6% on a year-to-date basis.
We've been adversely impacted because of domestic roaming, which is what we offer to TM, and also the inbound roaming, which has come down, as well as the interconnect, where the rates have been revised in the beginning of the year, and our conscious decision to reduce the hubbing revenue, which was at a very low margin of aggregated margin. That's having an impact on the overall mobile service revenue for us. However, a positive development of the cash flow as well as improvement in profit. Just to clarify, the last year, when we look at profit, we also include some of the voluntary separation costs, which we did in last year. If I go to XL, I think it's fair, a real strong performance, continuing 8 quarters of improved EBITDA quarter-on-quarter. Extremely good compared to how the market development is happening.
Not only on the top-line growth, but also improvement in the EBITDA margin by nearly 19%, and improvement in free cash flow, as well as recording 4 quarters of profit in Indonesia. Extremely good performance for us in Indonesia. If I go to the next one, Robi. Extremely good performance when it comes to revenue, a double-digit 10.4% revenue growth on a year-to-date basis, better than the market. Improved free cash flow and improved profit.
Here you would see a negative impact, largely because last year Robi did sell some towers to our own sister company, EDOTCO, which did result in one-off gain, which if I normalize for two factors, one is the gain last year from the sale of towers, as well as normalizing for the tax implication, which has come because of the new taxes introduced this year, we would see around close to MYR 5.5 billion improvement in the profits of Robi. Extremely good performance with EBITDA margin improvement of nearly 68% year-on-year. I don't intend going through all the other operations given the time and the opportunity to give much more time for the Q&A. Fair to say all of those, with the exception Celcom, which is largely driven because of the market condition as well as the lowering of the ILD have done fairly well.
What I will do now is to take you to slide number 21, which basically talks about where we are versus our KPI or the headline numbers which we had communicated at the beginning of the year. For revenue growth, we would be marginally lower, and that's because of conscious decision to hold them on device revenue in terms of our markets, which has an impact on the overall revenue. Revenue excluding devices, we should be above what was our target. EBITDA, strong EBITDA. We should be heading closer to a double-digit growth on EBITDA, and return on invested capital, we should be above 6%. CapEx, based on our conscious capital allocation strategy, we should be below our targeted CapEx for 2019. I think here, I mean, these are fairly well-known risks.
For example, I think we do see challenges of the unfavorable regulatory environment in Nepal with respect to capital gain tax. We recently got the Supreme Court judgment on the audit, better than what was expected. There are some open issues on that, and I think we continue to pursue the international forum for getting the right kind of decision on that. Issues related to Bangladesh, mostly coming out of the BTRC audit, where we have seen constant restrictions on getting new equipments as well as launching new products. There is challenge in these markets given the regulatory environment. Market conditions in Malaysia and Nepal remains difficult.
For Helios business, we are curtailing and to manage the OpEx there, where the intensity of competition in the Malaysian FinTech space is increasing, but it does help in terms of overall adoption of the e-commerce and services. Opportunities. Indonesia and Bangladesh continues to do well from an operations standpoint. Infrastructure, we are, from our EDOTCO getting new synergies and also looking at new markets, which should give us opportunities. Continued focus on structural cost base. We will get into more detail when we have the next week investor forum on how we are looking at structural costs going forward. Is there anything from Tan Sri?
All right. Thank you, Vivek. I think we just go straight to Q&A.
Thank you. We will now begin the question -and- answer session. If you wish to ask a question, please press star one on your telephone and wait for your name to be announced. If you wish to cancel your request, please press the pound or hash key, followed by the 2. Our first question today comes from Prem Jearajasingam from Macquarie. Please go ahead.
Hi. Thank you for the opportunity. A couple questions from me, firstly, on Celcom. If I were to look at the trend in your prepaid revenues, 1Q to 3Q, it looks like you are on an interesting glide path here, at least stabilizing on that front. If I look at 3Q versus 2Q, there's been quite a drop. Is it just pure seasonality or is there something that worked in 2Q which didn't work in 3Q? I suppose, the long-term question of how long before our margins start gravitating towards where our peers are sitting at the 40% level at Celcom. Second one is largely around 5G and infrastructure sharing.
I note the agreement between Maxis and Celcom on potential 5G network sharing, but could you help us understand what your philosophies are around 5G and how that network rollout is likely to take place in the Malaysian context? Thank you.
Thanks, Prem. I'll pass to Iran for his question, although I did hope that you would ask a question on EDOTCO since the CEO is here and performing so good.
Hi. Thank you for the question. I think the first question is about the prepaid Q3 versus Q2. Yes, there has been a slight decline to Q3 from Q2. Yes, we do see a lot more in terms of the prepaid to postpaid migration. I think this trend is coming more and more. I think this is something that we are working on. Hopefully with this, the trend and also what we're doing in the postpaid side, that we can actually capture more of this pre to post migration in the fourth quarter and also in the first quarter. This is a trend that we are handling internally. Your question about the philosophy around the 5G and also the sharing of the infrastructure with 5G.
There's a few things around the implementation of 5G that is being worked out within the industry in Malaysia, which has to do with the rollout of 5G in Zone one, Zone two, Zone 3, and Zone 4. A lot of this is being discussed at the industry level and there's also a lot of infrastructure sharing initiatives that will take place, especially in the Zone 3 and Zone 4. I think this is something that will be unveiled soon by the industry. It's premature for me to speak about it at this point in time. The agreement or the MOU that we have with AXIS is for the purpose of sharing within Zone one and Zone two, which is more of the city centers and also the urban areas.
At this point, at the early stage where we are sitting down together to start thinking of planning what if we were to share the 5G network. For us, it does make sense because we know that this is a trend that's happening globally. 5G is being planned to be shared by most of the major operators in the world. If you look at the networks in Malaysia, our networks are quite comparable. Our 2G, 3G, and 4G networks are quite comparable. We think we are in a position of strength when the two of us are thinking about how to share the 5G networks together. As I mentioned, it's still very in the early stage, and within the next 6 months, we hope to be able to define further how the sharing is going to take place.
Just one very quick comment before I pass back to you, is that from a broader philosophical, I guess, in Malaysia Act, we know that 5G, in terms of the commercial usage, is still hard at this point in time to justify the incremental revenue and our ARPU and the cost versus the benefits is quite questionable. At the same time, we do understand, we do support wholeheartedly the government's intention to bring about leadership in the technology using 5G as the basis, and we take pride in that. The sharing is actually one way of doing it so that we can achieve both at the same time. I think AXIS probably has the same view. From industry perspective, we believe sharing is the only way to achieve the above objectives.
I completely concur. I was just wondering how two mobile operators with limited fiber capacity would undertake this task, especially when the network's going to be requiring a lot more fiber. There's going to be a fair amount of investment, especially given that revenue upside is probably pretty limited at this point in time.
Well, when we talk about fiber, yes, we do recognize that. Both parties have quite a reasonable amount of fiber, actually. If you look at our own fiber, our sites, the direct fiber, we have more than 30% of our fiber. If you look at in terms of collected sites that are already fiberized, we're talking more than 90% of our collected sites are fiberized. We have our sites, more than 70% are actually single hop to fiber. I think that's one point. I think together, if you look at what fiber that Maxis have, we could complement with each other. Through the sharing, I think we can actually achieve more together.
Yeah. If I can add a bit on that. Don't forget that on fiber, we also are working with Digi to build fiber, and we have a reasonably good deal with TM as part of our overall deal, the bigger deal with TM on fiber. While it can be from a cost-effective concern, but I think we have taken a lot of measures to ensure that we can be economical on fiber.
All right. Thank you very much, and good luck on that.
Thanks.
Thank you. We now move on to our next question, which comes from Choong Chen Foong from CIMB. Please go ahead.
Hi. Thanks for the call. A question for Robi and for Celcom. Starting off with Robi, I noticed a very big jump in the PATMI Q on Q. Could you just sort of help us understand what drove the direct costs, the sales and marketing, and also the staff costs lower on a Q on Q basis? We also saw that the interest costs are more than half on a Q on Q basis. Any color on that as well. Lastly, for Robi, the tax, if I looked at it's still at about 7.6% of the revenue. Can you help me reconcile that with the 2% turnover tax? Those are three questions on Robi.
For Celcom, you mentioned earlier about the free to post migration, that you're seeing a bit more into the third quarter. If I look at the full payment adds, that was down Q on Q. Can you sort of give us more color as to what happened on the full paid type? Was this due to more competition in the lower end segment? If you have some update on the competition front, it'll be quite useful. Same on the network cost, that rose Q on Q. Was there any one-off still in terms of the LTE expansion cost in the quarter? Do you see room to significantly bring this cost down or do you think this is the top right of the price and quality of network that you want to run for Celcom? Those are my questions. Thank you.
Okay. Yeah. I'll take the first one, then I'll ask Jennifer Wong Fen Chui to cover the second one. Robi quarter-on-quarter improvement in PATMI has come from a few factors. One is the continuing growth in revenue, which is great. Second thing is consciously, with a lot of focus around driving profitability, the overall gross additions in the market have been slowed down by Robi. They've been really focusing on quality of customers, versus driving through higher gross addition and then leading into those customers. There's been a reduction in the sales and marketing cost as well as the A&P cost, and there's been continuing effort on improving the network cost. One thing I must say about Robi, if you look at last full year since we acquired most of Robi, the cost in that company has remained flat.
Despite the fact that we've been increasing the network coverage as well as growing revenue quite consistently. Robi managing their cost structure extremely good. In addition to that, this quarter did help because their focus was lot more on quality of sales, which means they did lower down the gross additions or had overall negative impact on the net adds. We saw our core improvement happening in Bangladesh. I think that's the one issue. Second issue which you raised was why tax is 7.6% of revenue, whereas we talked about 1.75%. There are two factors. One is that the tax included the effect was from 2018. There is MYR 45 million negative impact only in 2018, which we have to account for in 2019, because this effect started with the accounting year 2018.
Second is that it's a body regulation in Bangladesh. The regulation is you have to pay the highest of either 3% of revenue or 45% of the profit before tax or advance tax you pay. In this case, they end up paying 3% because Robi's profits are very low, profit before tax. However, these are exclusive. These are not inclusive provisions, which means they do need to still continue looking at accounting of tax liability on the standard 45% of profit before tax, which means because they are making losses, they end up accounting for that as a deferred tax or deferred tax adjustment. There are dual impacts which are coming into the accounts of Robi . I think one at a point in time when they start making better profits, we will start seeing the adjustments coming in on account of the deferred tax.
That's the reason why you see a much higher percentage of revenue as tax versus the 3%. Thank you.
I'll take the first part of the Celcom question on the free to post and then apart from the Gen network cost. Yes, you're right. There's competition on the low-end segment is intensified and possibly today even on the postpaid. In Malaysia, you can get for as low as MYR 28, you can get a postpaid package. The result that we have seen, yes, we were doing free to post, but we focus more on the high-quality customers. If you look at it, even though our subscriber base has declined a bit on the postpaid, but our postpaid revenue has actually increased. We are attracting the low-end segment, the competition at the low-end segment through different means, not directly through our brand, but it's something that we are doing actively.
We are also working on also an alternative plan around how to address this low-end segment better in the fourth and also the first quarter of next year. I will pass on to Jen to talk about the network cost.
Okay. The network cost, I wouldn't say that there's been a lot of the network cost in Q3. Actually, the absolute amount is nearly almost flat. We haven't seen much of a result in terms of the cost savings because we need to do more. There's more efforts to actually to really achieve the results in the longer run. We need to take actions. One of the key things that we're actually looking at is to optimize the equipment that we actually have on site. For us to even do that, we need to do more investment before we can actually see the cost reduction in the longer run. I hope that answered the question.
Yeah. Can I just follow up on the network cost? Should we expect that the network cost would go up a little bit more from here because we probably need to make some investments first before we start to see some change coming through? Or are we sort of at the peak and, going forward, maybe 12, 18 months, we should start to see some benefits?
We are hoping that the number that we are keeping now is the peak. We shouldn't see it fluctuate very much over the next, in the short term. Over a longer period, we should see the amount actually decline.
Okay. All right. Got it. Thank you so much.
Thank you.
Thank you. We now move on to a question from Alex Goh from AmBank. Please go ahead.
Thanks for the opportunity. I got three questions. One's on your slide 21, where you showed us your expectation for our guide in terms of the KPI guidance. It's indicating that the revenue growth guidance that you took will come below, but the EBITDA and ROIC is going to be above. Does that mean you are expecting a lower cost to contribute to this higher-than-guided EBITDA growth? Where is that reduction in cost likely to come from? Is it from the network cost? Which region? Is it from, say Malaysia, or is it any of the other countries that have had like excess? That's my first question. My second question is regarding your Celcom, where you've indicated that it was impacting the wholesale revenue. I expect that was coming due to the impact on fiber broadband.
What was the impact to the net level for Celcom, given the fact that you are also taking the fiber from TM as well, as [inaudible] was sharing. What was the impact on a net level from this wholesale reduction? My third question is regarding your CapEx. You've indicated that it is likely to be lower, but going into next year under MSP requirements, do you expect that to accelerate? Even if you're deciding to be below this year, but in fact, it's going to be an off-year next year. Thank you.
Let me answer the first question before I pass to Prithiv. Just a clarification. The revenue here is the total revenue. If you exclude device, we do expect our numbers compared to the assumed revenue target to be higher.
We actually should have done X device. The device is actually fairly artificial. We can increase or reduce depending on our subsidy and our marketing programs. What is more important is revenue X device. Just to clarify. Having said that, we do expect EBITDA to increase higher than the increase in revenue. I'm sorry. Yeah, that's right. Because of the cost improvement. I think the device will give you a bit of color.
Yeah. Thanks. That's post clarification. As I said earlier, so far on year-to-date basis, revenue ex-devices are growing at 5%, right? While 2.4% is including devices. There's been a conscious effort to curtail devices where subsidies were high. That said, I think our focus on cost continues across operations. If you go back to the slide number 12, if you have that in front of you, we try to explain where are the cost savings coming from. Majority of the cost is coming from the network, IT cost, and sales and marketing costs. These are coming from essentially markets in Bangladesh and Indonesia, and also on sales and marketing somewhat from Celcom. In addition to that, we've seen some improvement on the soft costs coming from Celcom and Dialog. These are three broad heads where we see the savings.
As you recall, I think we gave a MYR 5 billion target of cost savings, and this year target was MYR 1 billion. From that perspective, we are already around MYR 800 million plus on the cost savings for this year. Yes, we lose it also by some of the expansion related new costs which we continue to incur in markets where there is still growth opportunity. That's where we are looking at costs. Broadly, should remain flat or very low single-digit increasing costs compared to last year.
Yes. [inaudible]
I think the wholesale revenue that we have here is actually the domestic roaming arrangement that we still have with TM. That's the main chunk of the wholesale revenue that we have. In addition to the hubbing revenue, which I think Vivek mentioned earlier, that we consciously are reducing the arrangement for hubbing revenue, because at the end of the day, hubbing revenue, the margin is actually very thin. That is one major portion of the wholesale revenue. On the other hand, the wholesale revenue from the domestic roaming arrangements that we have declined year to date as compared to YTD 2018. The main reason for the drop was mainly because of the renegotiation of the rates. The impact from the reduction is about less than 3% of the total revenue year to date.
Question three.
Okay. Question three. Yes, on the CapEx, I think in general, we'll give the guidance for the CapEx for 2020 and growth sometimes in February. When it comes to NFCP, yes, there is a lot of discussion still at the industry level. We foresee there'll be a lot more collaboration happening within the industry in order to achieve the NFCP targets. We hope by doing this, we could help to minimize the CapEx required to achieve the NFCP.
Okay. Thank you.
Can I just squeeze in one question regarding Ncell? Regarding the CGT situation, how much have you provided for the CGT so far in your account, and could you just give a bit more color?
If you recall, I think we did mention this when we had the quarter four of the full year 2018, the earnings call. At that point in time, we did say that we believe there is sufficient provisions available on different tax matters. Some of them are into different stages of litigation across the group. We think at this point in time, this was sufficient where we were, to take care of the liabilities of CGT. However, until we just get clear clarity or we come to a conclusion, it's very difficult to say what will the actual amount be. From our perspective at this point in time, we do see that overall provisions we have on various tax matters should be sufficient or should be sufficient as the situation stands today.
In this year, do you see any additional provision, anything for any other reasons?
At this point in time, I think we've also mentioned in our quarter release. That we still believe that the strength of our international arbitration position on that is very strong. Unless we decide to take certain other measures, we do not see any major implications, at least at the time. The tribunal proceedings are continuing.
Okay. Thank you so much.
Thank you.
Thank you. Ranjan Sharma from J.P. Morgan has our next question. Please go ahead.
Hi. Good evening, and thank you for the call. A couple of questions from my side. Firstly, on EDOTCO, I understand that Indonesia is thinking about revising the negative list, which might open the tower industry to foreign investments. Is that something that's on the radar for you? Has the existing data set of towers within the negative list is potentially revised? Also, Nida, how quickly can you explain what has driven the big improvement in EBITDA margin on the first quarter from the third quarter? You said ADA is breakeven now. Is that expected to remain breakeven in 2020 as well? Are we still on track for breakeven for Apigate next year, according to the initial plan? The last question is on TM. I understand that the domestic roaming agreement was revised down. I think it's a MYR 15 million loss in revenues for quarter.
Can you just share with us or remind us, what is that on Telco? Actually, I had a question to turn. Is it a driver for access to fiber? If we can share what the benefit for Telco was. Thank you.
Okay. Suresh.
Okay. This is Suresh from EDOTCO here. I think, at EDOTCO, we always look at countries that are within Asia, but also available and addressable so that we can own and operate the assets. If Indonesia does indeed seriously rethink the negative list, it would certainly be something that we would consider. Naturally, any entry to Indonesia would depend on the individual benefit attractiveness of what targets are available. It's possible, but I think we have to wait and see. On the XL question, I think I'll leave that to a bit an answer the question on the margin. I'll send that back to [inaudible]. Right. Just on the margins, and I'm just going to use the pre-MFRS comparisons here.
The main drivers, if you look year-over-year or successive quarters, is largely because of, I think, increasing revenue coming both from towers and I think a consistent co-location maintenance. We've maintained co-location ratios around 1.6, even though we continue to grow towers both from small asset purchases and organic growth. We've improved a lot of revenue assurance. I think there's a little bit of benefit from better billing over the last few quarters. I think more importantly also on better cost management, which has come a little bit from some closure of projects on relocation of some towers, which we have carried some higher costs last year. I think this year, lower maintenance cost overall.
Some of that is actually coming from Malaysia because we took over the Telco view for the group that we announced last year, OSS, and that actually helped us drive a lot more cost synergies as well. Those are the key items that explain our numbers.
Thank you, Suresh. Adlan, are you on the mic?
Yes, perfect. I think if you look at the tower being removed from the negative list, I think this has been talked about for the last three, four years. Unfortunately, it has not happened. I think we know that it's coming back again. However, there's no certainty that it will come anytime soon. As far as we are concerned, we have actually announced that we are running the tower divestment. I think that is expected to be completed by end of Q1 next year.
Okay. On question number three, before I pass to Telco back on the TM. The answer is that we are expecting them to be basically neutral by the end of this year for ADA, for the whole year actually. Therefore, we should expect some profit next year for ADA. For Apigate, the plan was supposed to be profitable next year. We see it as a bit of a challenge. Still in the plan, but a bit of challenge, and we hope we can still make it. Question number four.
Question number four in terms of TM. Yes, there will be revenue contribution from the government revenue has actually declined, mainly because of the fact that they are also rolling out their own access for 4G. Which is one of the reasons the government has actually declined. The arrangement with TM, there are three things that in the old arrangement. Which is one is this domestic roaming, which has roaming on that, on 2G and 3G only. The second one is actually on the fiber build out, which they have actually, in the old arrangement, we have got a favorable arrangement, favorable rate in terms of the fiber leasing from TM, and that is still continuing. The third one is on the HSBB.
The solution that we have in hand. Yes, it may have some solution dealing with a lot lower because of the end contract. Even with the second solution that we have, we've done the modeling thing. That is already a very concrete thing.
All right.
Okay, got it. Thank you so much.
Thank you.
Thank you. As a reminder, to ask a question today, please press star one on your telephone keypad. We now move on to Arthur Pineda from Citi for our next question. Please go ahead.
Hi. Thanks for the opportunity. Just two questions, please. First, I'm curious on your decision to partner with Maxis on 5G deployment. Were you not too busy in the discussion considering you were actually thinking about merging with Telenor Asia assets? Is it limited to Maxis from their rules or a two-way deal? Second question I had was just regards to your thoughts on the stakes that you hold across your subsidiaries. Are you comfortable with the ownership levels? Do you think this is ideal, or would you consider increasing or decreasing stakes in these various assets? Thank you.
Let me take the question about the arrangement with Maxis. Well, as you know, the current 5G technology is still standalone. You have to take a few things into account. I think I'd like to try and say that this is actually the very early stage of both planning how to share the 5G network in zone one and zone two. We're starting with a non-standalone, which is depending very much on the 4G network that we have. At this stage, one of the reasons that we're working on with Maxis first is, one, in terms of the compatibility of the network, and we have to build the network size. Number two, the more important one is because we are using the same vendor for our 4G network.
This makes things a lot easier to plan and also to look at the migration from 4G and how to put the 5G on top of it. I think that's where we are. There is a bigger arrangement that we're looking at in 2024, which involves more players and more operators. That is something that we're not ready to discuss yet openly.
On the second question, to be frank, it's a bit tempting to actually increase our ownership in a few of our opcos. What we have, at this point, directionally, decided that we want to preserve our cash. We believe that there might be a need when it comes to consolidation or given the assets in the market.
How about the other way? Would you be willing to maybe reduce and actually monetize some of the assets?
There's nothing imminent at this point in time. If we were to do that, it's more not because of a belief in the outcome, but we want to use cash for maybe there's an opportunity for consolidation again. It's quite possible, but nothing imminent either way, to increase or to decrease.
Thank you.
Thank you very much. Yeah. Just to add to that, at the same time, we are actually looking at potential strategic investors in some of our possible consolidation that we want to do or in new areas that perhaps that we might embark on.
Okay. Thank you.
Thank you. We now have a question from Piyush Choudhary from HSBC. Please go ahead.
Hi. Good evening, thanks for the opportunity. Two questions. Firstly, at Celcom, could you, on a high level, elaborate on the outlook for the EBITDA margin for next year? We've seen improvement this year, what's the outlook for 2020, if any? On the regulatory side in Malaysia, any timelines for the spectrum allotment that you could share? Lastly, on EDOTCO, organically, what's the outlook for the tower growth for next year, and which countries would likely drive that? Thank you.
Okay. Let me pass this to [Suresh] since he's really into the working from practice.
Hi. It's [Suresh here] again. On the organic outlook, we would still expect that as a maybe slower growth than this year, but close to double-digit growth. That's in terms of number of towers or at least tenants. The mix will actually probably shift a little bit. We expect that Malaysia will have a little bit more of a gestation period waiting for 5G, so there may be slower growth here. We anticipate higher growth in the newer markets that we're in. Pakistan, we've recently almost completed an entry into the Philippines as well. We expect that some countries like Myanmar and Bangladesh also will have some reasonable growth. Maybe a little bit away from Malaysia and coming from newer and more emerging markets.
Okay. Super. EDOTCO revenue growth of lower double digits is not good enough. Celcom? With EBITDA?
Yeah. On the EBITDA margin. I think with all the initiatives that we're taking, especially on the cost management as what was mentioned before, we hope to see that the current EBITDA margin will continue the upward trend. Hopefully, to be around, today we are about 38.5%, to be close to around 40% in the next year or so. On the regulatory front, the bracket for the spectrum, I think there are three bands being looked at, the 700, 2300, and 2600. The latest indication of the process by the regulators is to be around the second quarter of 2020.
Thank you.
Thanks a lot.
Thank you. A quick reminder that it's star one for your questions today. There are no further questions at this time. We will now pass the call back to Tan Sri. Please continue, sir.
Thank you very much. I just want to end this again with a thank you to all those participating in the third quarter 2019 results. I will talk to you next quarter. Thank you.
Thank you very much.
Thank you.
This concludes today's call. Thank you for your participation. You may now disconnect.