Singapore Telecommunications Limited (Z74) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to the Singtel's Half Year FY '21 Results Conference Call. [Operator Instructions] Ms. Sin, over to you.

Thank you. A warm welcome to all investors and analysts. You are listening to Singtel's earnings conference call for the half year ended September 30, 2020. My name is Sin Yang Fong, and let me introduce management on the call. We have Mr. Chua Sock Koong, Group CEO; Mr. Yuen Kuan Moon, Group CEO Decimate; Ms. Kelly Bayer Rosmarin, CEO Optus; Mr. Bill Chang, CEO, Group Enterprise; Mr. Samba Natarajan, CEO, Group Digital L!fe; Mr. Arthur Lang, CEO, International; Ms. Lim Cheng Cheng, Group CFO; Ms. Jeann Low, Group Chief Corporate Officer; Mr. Quah Kung Yang, CFO Optus. Before we start taking questions, I would like to invite Sock Koong to share some highlights from this set of results.

Thanks, Yang Fong, and good morning, everyone. Thanks for joining us for Singtel's financial results for the half year ended September 30, 2020. I'll just share with you some key highlights from our business. The group performance for the half year was weighed down by the fixed business in Australia, and that was exacerbated by the COVID pandemic. Lower equipment sales and margins, ongoing price erosion in Carriage combined with adverse margin impact from the higher mix of NBN resales and declining NBN migration revenue contributed to the sharp decline in consumer [ of ] service revenue and earnings. COVID-19 [ excess basis ] declined as roaming revenues fell sharply due to travel restrictions and border closures. While economic uncertainties impacted demand, we delivered a strong performance in ICT services. Our digital investments set us up strongly for growth, notwithstanding short-term disruption in demand and supply conditions. NCS recorded strong growth across its line of business. We also saw good momentum in cloud and cybersecurity services in Asia Pac. We have been unrelenting in our support to customers and the community during this difficult period. More than ever, the pandemic has highlighted the importance of the critical infrastructure, seamless connectivity and digital solutions that we provide. We are accelerating our assets to position ourselves for the digital economy. Digitalization remains central as we move customers to our digital channels and platforms and improve our operating model and processes. We are also making significant investments into our 5G network and capabilities, which will unlock new revenue streams and deliver returns over the mid to long term. For the half year, revenue decreased by 10% due mainly to fall in equipment sales, roaming and prepaid mobile services. Optus revenue and earnings declined were marked by headwinds, particularly in the fixed business and intense competition. We also experienced delays in deferrals in ICT and infrastructure spending with the rollout projects disrupted by manpower constraints and site access. Despite the disruption, ICT services grew 8% and with robust performance from NCS as it delivered strongly on government projects. EBITDA fell 19%, while EBIT declined 44% on higher depreciation and amortization. Regional associates' pretax profits rose 11%. Airtel delivered stronger results led by operating momentum in India and Africa, offsetting lower contribution from telecom sales due to aggressive price competition and a prolonged COVID lockdown. Underlying net profit declined 36%. The group recorded a net profit compared to a net loss last year as there were higher exceptional losses booked last year related to provisions for the India adjusted gross revenue regulatory method. Also just want to highlight that on the sequential quarter basis, we saw improvements in our businesses in the second quarter as lockdown measures ease and customer spending gradually return. Carriage services posted sequential quarter growth, although they have not reached pre-pandemic levels yet. In Singapore, mobile service revenue growth was underpinned by increased prepaid usage, as previously quarantined foreign workers return to work progressively. Australia consumer mobile service revenue rebounded in the September [ 4th ] [Technical Difficulty] As posted, in the enterprise segment, Carriage and ICT revenues rose steadily, aided by business reopening and also resumption of projects. Amobee also posted revenue growth in the September quarter, led by a modest recovery in ad spending in North America. Let me quickly just wrap up by talking through of the outlook and dividend. I think in view of the continued uncertainty in the economic environment, we will not provide guidance on the outlook for the year, except that dividends from the regional associates will be approximately $1.3 billion and net group capital expenditure, including 5G network, will be around $2.2 billion, comprising $1.5 billion for Optus and $700 million [ U.S.A. dollars ] and $700 million for the rest of the group. The group has also -- the Board has approved an interim dividend of $0.051 per share. That represents approximately 100% of underlying net profit for the half year. Total dividends for the full year are expected not to exceed the group's underlying net profit. We are committed to our investment-grade credit rating, and we'll review our dividend policy at the end of the financial year when there's more clarity on the impact of COVID-19 pandemic on the group's businesses. So with that, let me hand it back to Yang Fong.

Thank you. Thank you, Sock Koong. Participants, please be advised that this call is being recorded for playback and transcription. We will now invite questions from participants. Our operator will assist you to put through your questions.

[Operator Instructions] We have now the very first question in queue from Piyush Choudhary from HSBC Securities.

Two questions, please. Firstly, on asset monetizations. Could you give us an update on potential monetization of Australian towers or any other value unlocking initiatives, which group is pursuing? Secondly, on 5G in Australia. Can you tell us your aspirations for network coverage in terms of population coverage by end 2021 and 2022? And what level of CapEx spends would be required over the next 3 -- 2 to 3 years to achieve such targets?

Maybe I ask Cheng to lead off on the asset monetization question. And then Kelly, you want to talk about 5G in Australia. Piyush, I should just note that we have guided the CapEx spend for this year, but not for outside of this year.

Piyush, thanks for the question. I think on the asset monetization, we are making good progress. We're working through, obviously, the different regulatory, legal and the valuation of the assets. So as and when, when we are ready to launch, you'll definitely hear about that. So over the course of the last few quarters, I think analysts have also asked about other forms of monetization. We've always said that we will continue to look at any other noncore assets and if there's anything that needs to be monetized it's something that has always been on radar.

I think maybe we also have seen the update of the tower sale in Indonesia. Telkomsel had sold the first part of the towers in order to basically a telecom subsidiary. So that has already been completed, yes. Kelly, over to you on 5G.

Thanks. So we are excited about our 5G plans. As you may know, in the 4G realm, we've gotten to 98.5% population coverage. And we're now turning our investment to 5G. We also recognize that there's -- not all 5G is created equal, and we're very focused on not just the population coverage, but we are rolling out Australia's fastest 5G. So we do have some strong CapEx plans. We gave guidance for this year, and we will be increasing our coverage progressively over the next several years to make sure we bring the benefits and the real tangible benefits of speed, low latency to our customers across the country.

[Operator Instructions] Our next telephone question is from Pang Vitt from Goldman Sachs.

Two questions from me. Firstly, on Australia. Can you update us on the competitive landscape only for paid and ARPU coming in a bit weaker? So I want to get some color? And how would we expect the amaysim acquisition to impact your business as well? That's the first question. Second question is on the enterprise business. Can you give us an outlook on interpret spending when we are entering Phase III in Singapore.

Sorry, can you repeat the question? We didn't hear any of it.

Sorry. Can you hear me now?

Not very well.

Okay. Let me try again. So the first question is on Australia in terms of competitive landscape. Can you give us an update on how it is going over there? And is there any more color you can give us in terms of the acquisition of amaysim and how this would impact? And second question is on the enterprise business. As we are entering Phase III in Singapore, can we get some outlook on enterprise spending? And any update if you can give us?

Maybe we get -- Kelly, you can talk about the competitive landscape in Australia. And then we pass over to Bill to talk about the enterprise business outlook, particularly in Singapore.

Thanks. So we operate in a very competitive market in Australia. What we are expecting is strong competition from the MNOs but an increasing move towards market repair and monetizing 5G. We also have a very competitive Tier 2 market with a large number of MVNOs and that market has been growing its share in the Australian landscape, especially as we've stepped into the first recession in 29 years in this country. So you will have seen that we announced that we will be seeking to purchase amaysim. They are the largest player in the MVNO market worth more than 1 million customers. We think that, that will put us in a very strong position to compete across the full spectrum of the Australian market. We feel we've been performing really well in the Tier 1 segment of the market, and we now want to have more pieces covering the whole board. We also announced that we'll be launching our own native digital GOMO brand into the market. So we do feel we will be in a position to compete very fiercely across every aspect of the market.

This is Bill. On the enterprise trend, so if you think about what's happened in Q2 versus Q1 in our first half results, the Q2 performance was much better relative to Q1 because of the opening up of the locations and therefore, allowing us unlike the first 3 months, there was a complete lockdown circuit breaker in Singapore and Australia. And Q2, there's more activity, so we are able to continue to deploy our projects and therefore, the uplift. But more importantly is the long-term trends, what we saw in this whole COVID environment is in discussions with many enterprise leaders, the digitalization trends in enterprises have definitely picked up tremendously. And what we hear is 5 years of digitalization brought forward in a matter of months. And we believe this is going to be an ongoing trend so from the largest enterprises to smallest SMEs, everybody is embarking on this to be relevant in this whole digital economy. So we are focusing on our capabilities. Our ICT growth in the first half was 8%, stronger in the second quarter as more activity open up. And where we are seeing growth is in the areas around cloud services, data centers, carrying cloud companies, largely in cybersecurity as small companies build up their digital footprint. There's more -- no [ need to, have ] companies working from home and therefore, understanding the new sort of environment that they're operating in. So activities in that. And also in the areas around digital capabilities in our SI business, NCS, we set up an NCS next organization. That's the focus on digital capabilities. And today, our digital revenues are 41% of the overall mix of NCS revenues. So it's an area of clear focus for us to build up aggressively. We've also recently made an acquisition in Singapore specifically in the digital services area, a company called 2359 that will help to extend and complement our capabilities that we're building organically. So a lot of focus has been pushing down this. Obviously, IoT and 5G, we're preparing for the enterprise engagements beyond connectivity, ensuring that we have the means to help enterprises transition to these low latency applications, things that are robotics in factories, things that are like autonomous systems in ports, AR, VR, very high-intensity throughput and low latency applications in a number of these large enterprises. And so that we move from a pure connectivity play to one that aggregates connectivity, low latency, high throughput to the edge cloud. And also to the data storage at the edge and ultimately, looking at the data analytics and AI at the edge to deliver sub-5 millisecond kind of performance so that enterprises can really unlock more digitalization and speed that up into this next few years. So that's why we are focusing on. And the focus is not just on large enterprises. We want to make sure that our capabilities extend down to SMEs. We have certainly worked with the government agencies and our partners in offering a lot of the bundled solutions to help SMEs work from home at the start of the pandemic, the onset. And very quickly help them transition in not only providing all these unified comps and communications over fixed and mobile networks, but also in ensuring that we provide security packages. And so we work with the government agencies like Enterprise Singapore to really help drive the adoption. And so we've got a major campaign that we're pushing. It's -- let's get digital for the SMEs and providing a lot of handholding to SMEs to help them cross this digitalization chasm. So those are the key things we believe with what's happening. And the new normal will be more digital, and companies would be focusing on that. And that is essentially where we are meeting those needs and building those capabilities and capacity to address that. Thank you.

And next telephone questions from Eric Choi from UBS.

I just had 2 questions, both for Kelly, sorry. And both on the 5G opportunity in consumer. First question is 1 of your competitors, Vodafone has said 5G handset penetration might only reach 20% by the end of 2021. So just wondering if you think that might be a bit too conservative. What percentage of your gross adds are being put on at 5G handsets or with 5G iPhones today? And then secondly, there was a really interesting stat out of the Telstra Investor Day today. They said 72% of their iPhone preorders are on plans where they're implicitly charging for 5G. And that's resulting in a $55 to $60 average sign on ARPU for Telstra. So just in light of that data point, I just wanted to get your updated thoughts on monetizing 5G maybe even next year. It just seems like if you did something similar, even accounting for some discounted Telstra, that would be pretty accretive to your ARPUs?

Eric, thanks for your questions. I think before I answer them directly, I also want to remind you that we have the FWA product on 5G. So we are already monetizing it. And we're seeing very high levels of customer satisfaction. We've recently rolled out a 2-tiered approach to that product. So at home, you can get up to 100 megabits per second speed and an uncapped speed at a higher price point. So that's one of the ways in which we're really monetizing 5G and are well ahead of the rest of the market. When it comes to mobile, I think answering your second question first. We've been quite clear that our approach to monetizing 5G needs to take into account how ubiquitous 5G is, how many customers can access it most of the time. And also ensuring that there is true value add, i.e., demonstrably higher speeds on 5G than 4G. When we reach the point that we believe those conditions are met, we will, of course, seek to monetize 5G. And so I agree that there is upside potential baked into our approach. In terms of 5G phone penetration, I'm not sure we've done the same calculations as some of the others. What I can say is that the iPhone 5G has had a significant impact. A lot of customers are opting to upgrade to 5G forms. We feel we're at the forefront of that. We have market-leading price positioning. We're giving a flat 5G phone price match guarantee, and we are seeing strong demand for 5G phones at the moment. So I expect to see 5G phone penetration continue to climb over the next few years. And we're certainly putting ourselves in a position to benefit from that.

Next telephone question is from Prem from Macquarie Securities.

This is Prem Jearajasingam from Macquarie Securities. A couple of questions from me, please. Bill, I think in the commentary earlier, there was talk about the government element kicking in to support enterprise revenues. Could you give us some color as to what the split is between government and nongovernment revenues are and where this growth is really coming from? Has 2Q really been driven by government? And therefore, going forward, we should see the private sector kick in, in a bigger way, therefore, the growth at enterprise could actually be much greater? Or was there an element of catch-up in 2Q, which may have exaggerated the performance somewhat? That's number one. And number two, the power sale in Indonesia. This seems to have been something that was long coming, but it's finally happened. Could you give us some color as to how your views on the market structures have changed to get this done now rather than previously? And what does it mean for the future of Telkomsel?

Bill here. So my earlier comment on government was more of government being a partner and the activity there specifically referred to them through Enterprise Singapore, helping SMEs. And so in that context, we are working with them as they issue grants to help SMEs digitalize. And we being a technology and a services provider enables it. So offering SMEs packages to allow them to digitalize, unified comps, work from home. Cybersecurity. So those are the things we put up. So it is not meant to be government as a consumption of those services. So that's the part that I was describing. However, the government is a large part of the mix, especially with regards to NCS. And if you think about the NCS business declared in our MD&A, and we broke it out. It's about $2 billion a year and out of $6 billion. And you can see the numbers in Q2 at -- in the first half, NCS would be at about $1.055 billion, a significant part. More than 70% of that is related to government agencies, statutory ports engagements. So that's that, but they've also got a growing and a faster-growing part that is in the commercial market, that is also seeing digitalization trends and also regional market opportunities, which is also growing fast, in fact, faster than what we are seeing in the government sector. And that clearly is our plan to really diversify NCS beyond the government sector. So I think the government in that case -- in this situation for NCS largely is a major customer. At the same time, government, it's also a major partner when it comes to helping the industry and especially SMES, they get a lot of help. The other part is, obviously, in 5G, the government is very keen to support the industry in the transformation and 5G, and that's where they issue grants beyond the SMEs, for larger enterprises. And that's where we're also working with them. But that, to me, it's really helping the enterprises to adopt 5G. So those are the key things that we would work with the government agencies very closely as a partner and as a customer. Sock, do you want me to take the point on the towers?

I couldn't unmute. Yes, please.

Okay. Sure. Okay. I think on the question on the Telkomsel towers. So just to give you a sense of the deal. So Telkomsel is disposing or selling it's a portion of its towers, so a total of about 6,000 towers through the wholly-owned subsidiary of telecom, Mitratel, and we're doing it over 2 phases. I think the rationale for doing this is a few things. One is it's really the focus on value creation for Telkomsel as well as the shareholders. I think we the company did an extensive review of its tower assets and bearing in mind the value of these towers as well as the potential upside that is -- that can be gotten from some of these towers and also factoring in the ongoing business operations for Telkomsel. Meaning, as you know, once you sell these towers, there's always that these costs and the operating costs that would have to be factored in to lease back these towers and to use these towers. So all this were factored into the consideration to decide on which towers had to be disposed of, which towers could be disposed to Mitratel. So I think the focus is definitely on creating value based on its assets that is on their balance sheet now. And I think we all know tower assets today are trading at much higher multiples than the operators, given the market conditions in the low interest rate environment. So the Telkomsel team, both the management team and the directors felt it was the right time to do it. And of course, with the potential of changes that may come. Of course, that's something we still do not know, and we leave it to the government of Indonesia to make decision. But with the potential changes on the omnibus law, we felt there was quite timely to undertake such a transaction. So that's really the thinking behind it. And as a shareholder of Telkomsel, we support basically all our OpCos to consider ways where we can unlock value, particularly with the assets that each of our operators has, whether it's towers, whether it's fiber, whether it's various assets. I think this is something we will support and strongly support as a shareholder.

Perfect. That's very helpful. Could I just 1 follow-up on the Telkomsel issue. Would you having raised this money, look at injecting it into more digital businesses? Or do you think this money comes up to the intel level?

We are exploring all options. As really at the end of the day, as shareholders, we encourage the company to do -- to really focus on a few things. Number one, definitely to create shareholder value. Now shareholder value can be created in a very short term, i.e., just do a single pot. But I think what we are focused on is to create sustainable shareholder value. This could mean investing back into the core business. It could mean investing in digital businesses potentially, but it could also mean a dividend back to the shareholders. And this is something we're working with the company. We're exploring and of course, working with our partner, Telkom. We're exploring the various options. But also bear in mind, Telkomsel has a very strong balance sheet. So there is a possibility that for any funding for its core business and expansion can be funded through its own internally generated cash flows. As well as potentially growing. So actually, Telkomsel today has a lot of options. So we're working closely. And maybe in the next few weeks, the company will be working on it and then we make a decision on that.

Our next telephone question is from Ranjan Sharma from JPMorgan.

It's Ranjan from JPMorgan. Before I ask my questions, I just wanted to say good luck to Sock Koong and Moon. So for my 2 questions. Firstly, can you share your thoughts if you should expect any change in corporate strategy? Singtel share price is back to where it was in 2005. So is there any change in the company's direction to create value for shareholders? The second question is on 5G pricing strategy. I guess, it was just mentioned that 5G might not be priced at a premium initially, but the premium could come through at a later stage. So should we think that the ARPU inflation for 5G would only happen in the mid to long term rather in the near term? And that just sounds a bit counterintuitive as well because for most markets and for the 4G migration, it was initially seen, 4G was initially seemed to be priced at a premium and then the premium faded with competition. So just would like to hear thoughts on how you think 5G impacts ARPUs across your markets.

Okay. Thanks for your good wishes. I think we've got to give more a bit of time before he comes and talks about what his strategic thrust and initiatives will be. So I think let us give him a breather and ask the question in the new year because he takes over officially first Jan. So [ respectful ] we ask for your indulgence and patients for another month or so. Okay. I think on 5G pricing strategy, the way I would look at it is, I think about 5G is really how we position 5G rather than just a pricing strategy. If you look at 5G and it's being positioned [ TRV ] as connectivity, you are back to the same issue, okay, because everybody has significant network capacity. Yes, people will -- that's why initially, you may get a bit of a price premium because of higher speeds, very quickly because of the excess network capacity, the price premium get eroded away. That's not a great place to be. And we've all seen increased ARPUs coming from higher data usage. But we all know that when there's excess capacity, that kind of premium gets eroded very quickly. So that's why when we talk about 5G, it's always around how you position 5G, we -- and operators should not look at 5G as providing connectivity only. And that's why you've seen the group -- our group being very, very focused on enterprise use cases because that would be a new application of a new revenue flow for us. And if you look at the speed capacity, the latent, a very, very low latency, it maybe lends itself to applications and users that we can't do with 3G and 4G previously. So it's a very long -- it's a very long and not indirect answer to your question, but I would say that while you try and get as much premium as you can on the connectivity side, your big upside is really when you're able to build services and platforms beyond just pure connectivity.

Our next telephone question is from Vandana Luthra from HSBC.

I have 2 things that I wanted to touch upon. Firstly, on spectrum costs. So would you be able to give us some indication on what is the spectrum cost you plan to incur this year and also next year, both at the Singapore level and at the Optus level? And secondly, I wanted to ask you about the subscriber numbers for Optus. I see that September quarter, mobile subscriber numbers are down versus June quarter. Could you help us understand what are the dynamics here that's leading to a fall in subscriber numbers?

Okay. I think on [indiscernible] those spectrum that has already been secured, the cost has already been announced and also the premium details. So it will be as a change to quickly run through this. The IPO [indiscernible] the details on the spectrum that we have already won with the detailed payment study. Cheng, do you have the data, okay?

I think for the [indiscernible] side, the spectrum retail that was I shared was to the tune of about $55 million. So that has already been announced okay. I think for Australia, this year, we are not expecting to pay any more spectrum costs per se. All right. So any new tranche of spectrum coming on-site is only when it's announced and if we decide to compete for it?

You have a question on subscriber numbers. Maybe I'll pass over to Kelly.

Thanks for your question. I think the big change in subscriber numbers is mostly in prepaid subscribers, and it reflects that we have a very strong business in selling prepaid SIMs to travelers and inbound immigrants. And of course, due to the restrictions and various lockdowns and travel restrictions, that has not been a market that's been available to us. And so relative to the previous year where that market was thriving, there has been a reduction in subscriber numbers. So that accounts for the majority of that. There is also an impact that's happening as we see one of the MVNOs, who's actually a brand that's owned by the merged TPG Vodafone entity is now moving its subscribers to TPG, and that has a big impact that has started to play out and will continue to play out over the remainder of the year.

Our next telephone question is from Ian Martin from New Street Research.

Just interested in the huge increase in data usage, both Singapore, which went up, I think, 20% on the June quarter and Australia, which went up from like 10 gigabyte to nearly 14 in the quarter. And I presume that's mainly 5G related. Is it -- there's no fixed wireless in that? And I'm interested in how it affects network management and user management offer. This is such a big issue with some 5G users, they're switching off 5G rather than running up to their data limits. And so I just wonder, Optus has got -- as all the network operators have, some big high use of data plans and I think promotion with 500 gigabyte per month. Is that designed around providing some peace of mind for those early 5G users?

I'm happy to answer these questions. So I do think that part of the increase in usage is a reflection of the lockdown conditions that people were facing. And we've seen data usage increase across all the technologies, including NBN at home, 5G at home, 4G home and mobile. So just a lot more consumption of entertainment and online services as people were working, schooling from home and entertaining themselves and connecting with loved ones through utilizing greater amounts of data. I'm very proud of how the network has performed to support that uplift across the board, the network has held up well. We have prioritized making sure that our capacity and throughput is strong across the network to really support our customers during this very challenging year. In terms of the high plans that you mentioned that we're promoting, I will take the opportunity to point out that, that has been a response to some very aggressive discounting by a major competitor at their top end plan. And so we thought we would make sure that they weren't getting away with that in the market without some strong competition. We certainly didn't lead that one, but it is a very good offer.

Our next telephone question in queue from Sachin Mittal from DBS.

A couple of questions. Firstly, on Australia. How much of the -- I'm [ just thinking ] into red is because of COVID and how much is due to the competitive positioning or the competition in the sector? And a related question will be, I mean, how confident do you feel to turn around Australia given that there will be a drop of almost $300 million in NBN fee going forward? So what are your plans? Can you share on Austrian turnaround efforts, whether it's the top line or on the off-site? Secondly, could you share with us what are the -- what are the cost of a 5G base station related to a 4G base station in both the markets in Singapore and Australia? That's it.

Thanks for your question, Sachin. In terms of the fall in EBITDA in Australia, more than half of that actually comes from headwinds related to the shift to NBN, both through the margin compression as we have lower resale margins for NBN, and we now have 90% of our customer base migrated onto the NBN. And also because of the fall in NBN migration payments off the peak from last year. So those are significant impacts, but they are more than half of the fall that you see, which is something that we believe was predictable and that the market was expecting. On top of that, there's a big impact from the lower numbers of equipment sales. Which is in line with global trends of people holding on to their phones for longer. So again, something that we think is quite consistent with what the market would expect. And then the sort of remaining amount we think is really COVID related. So some -- it looks like a very big number, but I think the paths are very explicable. And I would say, when it comes down to it, the competitive environment is not what's driving that decline. And if it were not for the COVID impact, we would have actually seen our mobile service revenue improve. As we are seeing ARPU improvement as the penetration of our choice plans continues to increase across our customer base. As you mentioned, there is still more of a drop in those NBN payments to come through. And so again, we're planning around that. We have some very strong plans around removing cost from our fixed business. A lot of that cannot be front-loaded because we can only shut down our proprietary networks once all the customers are off those networks. And we're not through all of the migrations just yet. So it will take a few years for those cost improvements to happen, but they're very much in train. And so again, we are quite optimistic about our long-term outlook there.

Our next telephone question is from Roger Samuel...

We've not got a question yet. One second [indiscernible] question?

My apologies.

Yes. I think your question is on the cost of the 5G position relative to the 4G. I think first of all, I think the 5G deployment in most markets are relying on existing 4G network that is building on top of it. So most of the time, we are talking about radio that is now fully integrated between 3G, 4G and 5G. So in the expense, we are talking about the same equipment is an incremental site. But if you look at it from a perspective of cost per gigabyte it's definitely going to be a much lower cost per digital buy because you do have a lot more capacity using the 5G spectrum to deliver the type of network capacity that 5G gets providing high-speed and huge bandwidth. But obviously, if you're using much higher band, then you would need a lot more sites to compensate for the poorer penetration because of the higher spectrum frequencies. So in general, I would say that the cost will definitely come down on the equipment as the global scale of 5G rollout over time. And we now know that a lot of the new base stations are all 4G, 5G [ LTE ], as you roll out, and that would be the norm, even in emerging markets where we are building up new 4G base stations, they will be really for 5G.

So just a follow-up on this. I understand because there is a lot more base stations required for 5G. So they will be some stations which are completely 5G base stations from ground up, right? So they don't -- I mean, we just build a 5G base station. My question is if today, you are -- now you're building some 5G base stations from ground up without any 3G and 4G. What kind of relative costs are you seeing on today's pricing, which actually will come down as you said, in the future?

Well, I think if you're talking about those additional base stations are primarily for millimeter wave deployment. And this millimeter wave deployment is really for very specific enterprise use cases where it is localized. For example, you may be building a 5G network for a manufacturing plant, and therefore, you need a high-density base stations around the area of a industrial park or whether this is a port or airport. So in those deployment, you are actually building for a specific purpose with specific enterprise customers, and you would measure that cost of that investment versus the potential business revenue coming in from the enterprise customers. By and large, if you look at deploying it for general mass consumer usage, you will be relying on a mixed spectrum of your existing reuse of your 4G spectrum as well as some of the newer spectrum by the 3.5 gigahertz mega -- bigger business -- megahertz spectrum, which will be definitely relying on existing sites to minimize additional cost.

Our next telephone question is from Roger Samuel from Jefferies.

I've got 2 very quick questions on Australia. First one is around the mobile customers, and I understand the drop in prepaid mobile. But postpaid mobile customers also fell a little bit quarter-on-quarter. And I'm just wondering if you can tell us whether it's the postpaid brand, the Optus branded customers? Or would it something else? And the second question is just on fixed wireless access for FWA. And I just want to confirm where you recognize the revenue, whether it's in mobile or in fixed?

Thanks for your question, Roger. In terms of postpaid, there was a small drop in postpaid. Some of that, again, is the impact of immigrants, et cetera. But actually, the largest correlation there is customers who no longer need the service. And there's a very strong correlation to economic areas where they've been the biggest job losses in Australia as a result of the COVID shutdown. So there is this impact of people not having as many services, rationalizing in the household and tightening their belts, given the economic environment and what people have gone through in the last few months. And we think that, that's highly correlated to the small reduction on the postpaid side in addition to the immigrant business that is missing from this year's numbers. On FWA, I think that we're accounting for that in our mobile numbers.. Does that answer your questions, Roger?

Yes, yes, yes. Yes.

Our next telephone question is from Arthur Pineda from Citigroup.

I have 2 questions, please. Firstly, can you get any color on the consumer operating trends for Singapore and Australia going into the third quarter? I see that you've seen a significant improvement in 2Q versus 1Q. Is this improvement sustaining into the third quarter as well? Second question is on the dividend from associates. You mentioned $1.3 billion. Does this reflect any potential for special dividends? Or is this just business as usual? In PT Telkom's call, for instance, they mentioned special dividends most likely on Telkomsel. And third question I had on Digital L!fe. I noticed that you have swung into a positive EBITDA in 2Q. Were there any one-offs here? Or is this a new normal now for Digital L!fe with the expenses taken out and ad spend reviving?

Sir, your question -- your last question, I didn't get it. Can you repeat that again?

So the last question was on Digital L!fe. It swung into positive EBITDA in 2Q. I'm just wondering, are there any one-offs? Or is this now the new normal that we should see for digital lag, given that [ hook ] is now out of the equation? And ad spend has rebounded.

[indiscernible] changed to take your -- the question on the associates dividend and the GDL EBITDA numbers. And then I think both Moon and Kelly can round off with the operating trends for our consumer business.

Okay. With regards to the dividend, I think you should just project the business as usual because, as you rightly call it special, we contented special. So that's really up to the Board decision in the near future. With regard to your question, I think you asked some question on EBITDA, whether there's any one-off for 2Q. I would say that maybe the more one-off item would be jobs credit, I think we have also disclosed the jobs credit. I think for like [ 4 million ] or thereabout for the first half results. You should expect that taper down, okay, for Q3 significantly because of the -- what was announced by the government, I think 10% of the total reach going forward, okay? To end December. So with regard to Digital L!fe, I think you also asked whether there's any one-offs. The so-called one-offs would be the -- of course, the closure of [ hook ] so we stopped having to take the base losses of hook, per se, so that to be so caught one-off.

Arthur, we continue with your first question on operating trends on Q1 to Q2 and looking to Q3. First of all, I think the improvement on Q2 over Q1 is primarily because Q1, we do have a sort of a side lockdown period where a lot of the businesses and consumers were all staying at home and businesses with shutting down. So that -- if we -- at Q2 versus Q1 has got a bit of that momentum that people are coming back out after the lockdown period and we are seeing sales coming through a bit skewed, much lower than pre-COVID days. I think going into Q3, we are seeing some positive trend in terms of customers buying more phones because of the recent new phone launches. So that's a good sign. And together with that, they are also upgrading into 5G phones and 5G plans. So I think this is some positive lines that we are seeing, but we are still impacted by travel restriction. Roaming will still be impacted because we do not see traveling coming back anytime soon, both for business travel as well as consumer leisure travel. In addition, using Singapore, we also note that the foreign workers population and expatriates have been shrinking in terms of the total population size in Singapore, that also will have an impact both on the mobile and the [ effect on ] business for Singapore. So this will continue to be a bit more pressure on both our topline as well as the bottom line costs. Roaming service is still a very good margin. And in the segment of prepaid where foreign workers, we have the lion's share of that, and that also impacted the more really than anyone else. We will see some positive issues in terms of buying momentum from consumers. So we'll be watching this space very closely in the next 2 quarters.

As far as your question goes relating to Australia, I think we have a lot of really good initiatives for the second half, including the 5G iPhone season, which is upon us now that we think will be very good. Then Christmas, back-to-school, the launch of GOMO, our second brand, so lots to look forward to. However, the economic conditions and sentiment remains uncertain. In Australia, there's been a lot of stimulus put into the economy, especially through the job keeper, job seeker schemes, which come to an end towards the end of March. So there's a lot of uncertainty about whether there's more job losses and economic consequences to come and how that will impact on sentiment. It could go both ways. It could really dampen people's discretionary spending. There could also be an element of the lipstick effect where people can't travel, can't do things, so they treat themselves for new phones. So we have to see how that plays out over time. As Moon said, we don't expect that roaming will return in the second half of this year. And also the pressures on our business through the continued NBN migration and reducing NBN migration payments are even stronger in the second half than the first. So hopefully, that gives you a good overview of the different factors that will come into play.

Our next telephone question is from Mr. Paul Chew from Phillip Securities.

I have 3 questions. Just 2 on Australia and just on 1 on just the cost saving. Not sure if you're able to quantify, but maybe can you mention some of the types of costs that can be removed post NBN migration? Second question on Australia would be, could you just share how did the MVNOs turn more competitive than MNOs since that you actually we had to acquire an MVNO? Of course, I would have thought the MNOs should have no better cost structure on paper at least. My final question is just on costs. I see that's a cost savings line item. I'm just trying to understand what does it actually mean. Like why was it -- what is the description of this cost item? And you also have an indirect cost, which I saw it dropped in Singapore consumer. I'm just wondering how much is that is fixed cost? And can we get like operating leverage from this so-called cost item post-COVID?

I can start with the first 2 questions. So in terms of cost out for -- that we can take out once we finish the migration. A lot of those costs relate to us currently running 2 proprietary fixed networks in HFC and ULL networks. So once customers have all migrated off those, we can take steps to close down those networks, having migrated the customers across. So those are all the network and associated costs of running those different networks. There's also a significant cost that we have in the moment in operating an NBN migration concierge. We put that in place to make up for challenges in how NBN service was happening in our market. And so we wanted to make sure customers were well looked after. So we've been handholding customers through the process of migration with this concierge team. So once customers are migrated, we will look to remove that cost from our fixed portfolio. And then we have a lot of ongoing operational costs that we hope we will be able to streamline once we have a single network on NBN, reseller margins, and we're looking to digitalize more of our processes as we eliminate some of that extra complexity with selling multiple technology solutions. So we feel pretty confident there's a strong plan there. On your question with MVNOs, I think the cost structure is very different. MVNOs do not operate stores, they tend to have more online or call center-based models. They tend to operate in a low frills part of the market, where there isn't value-added benefits or as strong service that accompanies the solutions there. So they have tended to price in a different point in the market because they're delivering a different value proposition to customers. I thought the third question was for Singapore.

Yes. Okay. Let me take the first sailing. Yes, thanks. I think we didn't give any guidance this year for the cost savings. So whatever that we have now, unfortunately, what we are seeing, showing is only for the half year mark. But I think with regard to the direct and indirect costs, I think the direct cost, I think we have actually given a definition of content as to what constitute the direct cost, which I think you can see on Page 13, Page 14 of the Singapore consumer side predominantly, but that will give a good description as to what really compromise -- sorry, include that we include in our direct cost, which is really cost that's directly attributable to all the revenue earned per se. I think we wanted to let this speak to also show investors that previously those that's directly attributable. It does go up and down in tandem with regards to the revenue line. And the indirect cost, obviously, I think you did ask about the cost savings with regards to that. So as we continue to digitalize our business or all, can we talk about the NBA network as we continue to retire the older network, then you will see some movement in those indirect costs. And hopefully, that in the medium to longer-term will actually come down.

Our next telephone question from today comes from Varun Ahuja from Crédit Suisse.

I've got 2 questions. First, I just want to check on the GOMO brand strategy, right? So GOMO has launched in Australia, you launched in Philippines and Thailand. So -- and if I remember, you mentioned historically that it is a separate entity, then within the single group. So when you go to associate, is it still controlled by Singtel? Is it becomes a part of the associate? And hence, they pay you some royalty for the technology and all that stuff. So just wanted to understand that arrangement and also long term, how do you see it? If it's a separate entity, how are you looking at it? Because some of the competition is looking at going across various markets and achieve a more digital telco solution. So just wanted to see how Singtel is looking at that particular business? And lastly, Singtel over the last 3, 4 years have seen a lot of challenges, starting with competition in overall Singapore market, then you've got associates competition and with COVID coming across and then challenges on the fixed enterprise business, margin pressure and even in Australia. Now looking into the future, I don't want any guidance, but is it fair to assume that most of the headwinds are behind the company and incremental, you should see some improvement? Any qualitative comment would be helpful.

Varun, maybe let me tackle the question on GOMO and comment a bit on the general competitive environment in the industry. Firstly, I think we started GOMO about 1.5 years ago in Singapore as a digital-only brand to compete in the Tier 2 segment of the market. We find that using a Tier 1 telco, giving it an infrastructure and allow it to operate on a totally separate digital brand gives us the flexibility to address different segments of the market. And that has worked very well for Singapore. And with that idea, we exported the same idea into our associates market firstly, with Telkomsel, where they very quickly launched a very equivalent brand called Bayou, which is really positioned very similarly to GOMO and also a pure digital-only brand to address the Tier 2 market where we are seeing a lot more price competition in that area. And soon after that, you see that both AIS and Globe have also adopted the GOMO branding and use it to create their own digital-only brand. So it is not a separate company. GOMO is not a separate company. It is still owned by each of the OpCo. And obviously, we all adopt the same brand GOMO. The intent and the idea is to create a regional network under the same brand because we do believe that in a post COVID, when traveling comes back to a bit more normal type of traveling condition, this particular segment of consumers who use GOMO actually travels quite a [ fair bit ] and by offering them localized offer as they travel around the region will be a differentiation for this group of customers. And of course, more recently, Kelly announced that GOMO will be introduced in Australia and Optus slightly soon. We also look forward to the launch of GOMO in Australia as a digital brand that will address the Tier 2 market in Australia as well. So we are quite excited about this, although it's very early days, but we do see that this -- the customer profile of coming onto GOMO is actually very different from those who come on to the traditional Singtel brand, the Telkomsel brand or AIS or Globe brands. So as you can imagine, both -- all the associates, Telkomsel as Globe and Singtel. We are all market leaders. So therefore, having a digital brand like GOMO actually help to address a segment of the market which we normally would not be able to get into. On your other questions on competitiveness, competition in all the markets, including our associated markets, emerging markets. The telco industry has been under tremendous challenge, especially when we start to move from 3G to 4G and now from 4G to the beginning of migration into 5G. So at every transition of technology, you see that we have to also migrate some of our old revenue streams for example, in 3G case or 2G, 3G case it was a voice revenue. Moving onto data. For 3G times. And then when we go from 3G to 4G, it was really about growing data revenue in place of voice revenue. So that transition for a very developed market like Australia is already done. So you see very relatively low mobile voice revenue in the market. But for Singapore, we do still see significant voice revenue both on the prepaid and postpaid market. And we are still going through the transition where we see data revenue growing, but being pressed -- the growth is really pressed by the decline in voice traffic. And for all our emerging markets, you will also continue to see this migration of 3G to 4G network as well as the revenue moved from voice to data. And obviously, you see that this will take time, especially in a big country, big markets like Indonesia, India, Thailand, you see that the migration of -- to a pure data network is going to take time and it's going to require a lot more investment. And now we are looking into the future of the next cycle of technology refresh, which is 5G. And as [indiscernible] has earlier mentioned, if it is purely on a connectivity growth on revenue, there's going to be that potential is there, but it's not going to be able to fully mitigate against all the old revenue that is eroding away. So we do have to look at 5G as a platform to enable new businesses for the enterprise as well as the consumer, to look at what type of applications that we can provide on top of that 5G network that take advantage of the low latency and high bandwidth, high capacity 5G network. So it fits very well into the industry trend, not just in the telco industry, all industry of enterprise moving into cloud, moving into digitalization, moving to edge computing. So all this would require 5G as an underlying enabling technology them to transform. So we are optimistic in the mid to longer term, where we see some of this new revenue stream coming on board. And Bill has earlier on also mentioned about NCS being in the right place to capture some of this value that's created through the digitalization of various industry, both on the government sector as well as in the commercial sector. So it is going to be long to me longer-term or midterm when we see some new revenue streams that is going to be turned on because of 5G.

Thanks, Varun. Okay. In terms of a time check, I think we've actually gone past the hour. I mean from my perspective, we've had a very good discussion. We hope it has been useful to our audience as well. So I think we can tell there are certain callers who might still have questions. So if there are no further questions, please feel free to reach out to the Investor Relations team. A transcript of today's call will also be posted on our website on Friday. So with that, we'll talk to you again and [ see you ] next time in May. Thank you so much. Bye-bye.