Singtel CFO Arthur Koh discusses H1 profit growth, digital infrastructure monetization, and Starlink competition dynamics.
Summary
Singtel Group CFO Arthur Koh reports over 10% underlying profit growth in H1, upgrading full-year EBIT guidance to mid-to-high single digits. He details the company's tri-brand strategy in Singapore and the distinction between Nxera's data center leasing and REAI's GPU-as-a-service model. Koh addresses Starlink's threat, arguing satellite is complementary for rural areas but faces urban and regulatory hurdles. The episode also covers Singtel's forward-looking dividend guidance of 19.5 to 20 cents per share and the economic viability of satellite versus terrestrial networks in low-ARPU markets.
Episode sections
- 0:04 – 2:08
Introduction and H1 Results Overview
Adam introduces Arthur Koh, Singtel's Group CFO, noting their last meeting was two years prior. Arthur highlights that despite global market volatility, currency weakness, and regional competition, Singtel achieved over 10% growth in underlying profits and 9% EBIT growth in the first half of the financial year.
- 2:08 – 7:37
Guidance Upgrade and Segment Performance
Arthur announces an upgrade in EBIT guidance from low-to-mid single digits to mid-to-high single digits. He notes double-digit EBIT growth across key operating companies (OpCos) including Optus, DIXO (digital infrastructure), and NCS, while Singapore's STSG business saw slight EBIT decline but sequential revenue growth of 4%.
- 7:37 – 11:21
Singtel Singapore Strategy and Enterprise Growth
Arthur details the 'tri-brand strategy' in Singapore (SingTel, Gomo, and a value-conscious brand) to target different customer segments. He emphasizes that over half of revenues come from enterprise, leveraging owned submarine cables and fiber for connectivity, which is seeing improved interest.
- 11:21 – 18:35
Digital Infrastructure: Nxera and REAI
Arthur explains the distinction between Nxera (data center landlord) and REAI (GPU-as-a-service provider). He stresses a disciplined CapEx approach, requiring secured power/water and high-quality customer contracts before building. He notes DC TWAS is filling up rapidly and REAI has deployed 4MW of its 6MW capacity.
- 18:35 – 20:32
Nxera Monetization and Scale
Arthur discusses plans to monetize Nxera, potentially via an IPO in Singapore, the US, or other jurisdictions. He highlights the combined scale of Nxera and STT GDC, noting a current operating capacity of 900MW and contracted capacity of over 1.5GW, offering significant future revenue visibility.
- 20:32 – 23:11
AI Risk and Capital Discipline
Addressing AI bubble concerns, Arthur asserts demand is driven by broad enterprise adoption, not just hyperscalers. He outlines a risk-mitigated approach: securing high-quality contracts, ensuring power/water security, engaging local communities, and avoiding circular financing structures.
- 23:11 – 29:07
Starlink and Satellite Competition
Arthur addresses the threat of Starlink, arguing satellite technology is complementary for rural/remote areas but faces physical limitations in dense urban environments. He highlights regulatory hurdles, such as spectrum ownership restrictions in some countries, and the complexity of foreign operators entering local markets.
- 29:07 – 30:35
Satellite vs Terrestrial Mobile Economics
Speaker_2 compares US mobile ARPU ($50-60) with satellite costs (~$30), arguing disruption depends on whether satellite costs fall below local ARPU levels. In Southeast Asian markets like India and Indonesia, ARPU is ~$3, making satellite economics currently unviable compared to terrestrial networks.
- 30:35 – 34:13
Singapore Mobile Market Sustainability
The speaker argues that while low data prices benefit consumers, they threaten the financial sustainability of operators needed to fund critical infrastructure investments like cybersecurity and AI. He warns that prolonged losses could lead to underinvestment, posing national security risks.
- 34:13 – 38:25
Dividend Guidance and Outlook
Singtel provides forward-looking dividend guidance, targeting a full-year dividend growth of mid-to-high single digits (19.5 to 20 cents per share) compared to the previous year's 18.5 cents. The speaker emphasizes this predictability as a commitment to shareholders despite near-term headwinds like a strong Singapore dollar and competition.
Claims worth verifying
Underlying profits increased by over 10% in the first half of the financial year.
Unidentified speaker (speaker_2) ·1:47
How to verify: Singtel Annual Report or Half-Year Financial Results for the relevant financial year (likely FY2025 given the Nov 2026 release date and 'first half' context), specifically the 'Underlying Profit' line item. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
EBIT grew by 9% in the first half of the financial year.
Unidentified speaker (speaker_2) ·2:29
How to verify: Singtel Half-Year Financial Results, checking the EBIT growth rate compared to the prior year period. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceEBIT guidance was upgraded from low-to-mid single digits to mid-to-high single digits.
Unidentified speaker (speaker_2) ·2:29
How to verify: Singtel Investor Presentations or Press Releases issued after the H1 results announcement, comparing initial guidance with the revised guidance. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceSingapore's STSG business saw sequential revenue growth of about 4% over the last three quarters.
Unidentified speaker (speaker_2) ·3:48
How to verify: Singtel Segment Reporting in Half-Year Results, specifically the Singapore segment revenue trends quarter-over-quarter. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Optus achieved double-digit EBIT growth in the first half.
Unidentified speaker (speaker_2) ·2:48
How to verify: Singtel Half-Year Results, specifically the Optus segment EBIT performance. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceNxera (DC TWAS) was at 30-40% fill-up in H1, expected to reach 80% in H2 and 100% within 3-4 months.
Unidentified speaker (speaker_2) ·12:41
How to verify: Nxera or Singtel operational updates, press releases regarding DC TWAS occupancy rates. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
REAI has deployed 4 of its 6 megawatts of capacity, generating income for one to two months.
Unidentified speaker (speaker_2) ·13:20
How to verify: Singtel Investor Presentations or REAI operational reports detailing capacity deployment and revenue recognition timelines. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Singtel acquired STT GDC slightly more than a month ago.
Unidentified speaker (speaker_2) ·19:41
How to verify: Singtel Press Releases or SEC/SGX filings announcing the completion of the STT GDC acquisition, dated around October 2026. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceCombined operating capacity of Nxera and STT GDC is 900MW, with contracted capacity over 1.5GW.
Unidentified speaker (speaker_2) ·20:01
How to verify: Singtel Investor Presentation slides or press releases detailing the combined data center portfolio capacity post-STT GDC acquisition. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceUS telcos have an average ARPU of $50-$60 USD, while satellite cost per user is estimated at ~$30.
Unidentified speaker (speaker_2) ·29:15
How to verify: Check ARPU reports from major US carriers (Verizon, AT&T, T-Mobile) for 2026/2026; verify satellite data cost models via Starlink or SpaceX technical whitepapers. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
ARPU in India, Indonesia, and the Philippines is approximately $3 USD.
Unidentified speaker (speaker_2) ·29:46
How to verify: Check quarterly earnings reports of Indian telcos (Jio, Airtel) and Indonesian telcos (Telkomsel) for 2026/2026 ARPU figures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Singapore has the lowest mobile data prices in the world.
Unidentified speaker (speaker_2) ·31:38
How to verify: Consult international telecom price indices such as the ITU (International Telecommunication Union) Mobile Telephony Price Index or Numbeo data for 2026/2026. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Singtel paid 18.5 cents per share in dividends last year.
Unidentified speaker (speaker_2) ·35:32
How to verify: Check Singtel's FY2025 Annual Report or SGX filings for dividend per share (DPS) history. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceSingtel guides for full-year dividend growth of mid-to-high single digits, targeting 19.5 to 20 cents per share.
Unidentified speaker (speaker_2) ·35:12
How to verify: Check Singtel's FY2026 Interim or Full Year Results presentation slides for dividend guidance. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
The Simba and M1 merger was called off.
Unidentified speaker (speaker_2) ·37:25
How to verify: Check SGX announcements or news archives from 2026/2026 regarding the termination of the Simba Communications and M1 merger. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Singtel is the first company in the market to introduce the concept of a 'Value Realization Dividend' (VRD).
Unidentified speaker (speaker_2) ·34:52
How to verify: Check Singtel's investor presentations from 2026/2026 for the introduction of the VRD framework. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Do your own research
- Verify Singtel's H1 financial results against the official SGX filing for FY2025 H1 results to confirm profit and EBIT figures.
- Check Nxera's latest investor presentations or prospectus documents for details on the 900MW capacity and 1.5GW contracted capacity.
- Review regulatory filings in key markets (Singapore, Australia, India) for spectrum ownership restrictions affecting satellite operators.
- Analyze Singtel's dividend history and cash flow statements to assess the sustainability of the 19.5-20 cent per share guidance.
- Compare Starlink's pricing and coverage maps against Singtel's terrestrial network coverage in dense urban vs. rural areas.
- Investigate the financial health of competitors like M1 and Simba post-merger collapse to assess competitive pressure in Singapore.
- Check power and water utility contracts for Nxera's data centers to verify the 'powered land' bottleneck claims.
- Review ARPU data for Singtel's key markets (Singapore, Australia, India, Indonesia) to validate the satellite economics argument.
Full transcriptexpand
0:04Unidentified speaker (UNKNOWN): Alright,
0:05Adam: welcome back everyone. My name is Adam. I have Victor.
0:07Unidentified speaker (speaker_1): Hi everyone.
0:08Adam: And today we have a very, very special guest. He is the group CFO of SingTail. Arthur.
0:12Unidentified speaker (UNKNOWN): Hi.
0:12Adam: Welcome back. Thank you for being here. And thank you for having us. Oh thank you for
0:18Unidentified speaker (UNKNOWN): the video. Yes. Yes. Uh This is
0:22Adam: a very, very nice studio that you have over with here. me.
0:24Unidentified speaker (UNKNOWN): Thank you.
0:24Unidentified speaker (speaker_2): Thank you.
0:25Unidentified speaker (UNKNOWN): Uh
0:25Adam: The last time we had you on our channel. it was two years ago. Yeah,
0:28Unidentified speaker (speaker_2): time flights.
0:29Unidentified speaker (UNKNOWN): And you
0:30Unidentified speaker (speaker_1): are the first management that comes on the channel.
0:32Unidentified speaker (UNKNOWN): Oh,
0:32Unidentified speaker (speaker_2): is that right?
0:34Unidentified speaker (UNKNOWN): Yeah, then after
0:34Unidentified speaker (speaker_1): that
0:35Unidentified speaker (UNKNOWN): the actually I recall,
0:36Unidentified speaker (speaker_2): I recall.
0:36Unidentified speaker (UNKNOWN): Yeah, yeah, yeah. I remember that. October
0:38Adam: 2024.
0:38Unidentified speaker (speaker_2): Well, thank you for the support.
0:40Unidentified speaker (UNKNOWN): I think we were just starting to do it. Yeah, yeah.
0:44Adam: Today we're gonna have a l look talk about the latest updates for SingTel and the question whether um Starlink could actually threaten your business. I think that's a really interesting one. Okay. Uh but to bring back bring the viewers back to two years ago, at that point in time was you were talking about the strategic reset for SingTel. Yes. SingTel twenty eight. Yes. And then you just re released your first half results for this year. Yes. Uh pretty good set of results. You wanna tell us a bit more about that?
1:07Unidentified speaker (speaker_2): Sure. Well first of all thank you. I I can't believe it was two years. It seemed like a long time ago but there's so many things have happened. But uh well nice to reconnect again. So so I think we start off with our results. I think we had uh I would say a very respectable set of results, and if I may say so, it was quite a good set of results. Um and the reason why it's good is you just imagine how many, so many things have happened in the global markets in the last six months, and not only just the global markets, in the geographies that we operate operated. in. Whether it's the Hormuz crisis in Singapore, you had the M1 and the Simba merger being called off, which created uh excessive competition in the industry. We had the AI uh markets kind of valuations went up and then it went all the way down and it swings like a roller coaster. Uh, but I and you had our all the regional currencies you know weakening against the very strong Sing dollar. So despite all that, right, I think uh the response. respectable showing was that our underlying profits had increased by over uh by 10%.
2:08And uh and that's important because the underlying profit is from which our core dividend is based on. And equally important is really uh our EBIT has actually grown. I think we started the year a bit more cautious, right? Uh and we actually guided the market that our EBIT would grow by low to low. mid single digits. After the first six months of this financial year and looking forward into the next six months, I think we are confident enough to say that we're revising that guidance upwards, right? To now to mid to high single digits. I think that's a very strong testament. And our EBIT has also grown by quite, I think, 9%. And it's really an indication across the board we have. seen all our opcos actually firing up. Whether it's Optus, our Australian business, DICO, our digital infrastructure company business, and NCS all had double-digit e-bit growth. Optus, you know, we you know, despite the issues and the problems that we have faced
3:08in the last two years since we met, I think the team has really hunkered down, really focused on going back. to the basics, you know, really building resilience in our network and building customer trust. And at the same time, actually, the underlying performance has improved quite significantly. And you see, you know, double-digit e-bit growth in Optus. What is interesting is Singapore, right? Our STSG business uh did not register an eBay growth. In fact, the eBay was slightly down year on year. But I think what's very important to note is despite the excessive competition that we are seeing in the Singapore market, the sequential quarters, that means from the last three quarters, that means from beginning of this year to the first quarter to the second quarter, there's actually sequential revenue growth, right, of about 4%. That's very important because if you look at some of our competitors, and I won't mention who, you compare what they have registered in terms of revenue growth and profits relative to our business, it's a very respectable showing. And I'm actually proud to say that I'm very thankful to say that you know not only have the revenues grown, our pools have also grown, right? And uh and I think equally important is the customer net ads, especially in our main brand, we're using leveraging on AI
4:28to actually improve on churn. That means the number of customers that we lose has fallen quite dramatically. And we can talk about it later. But I would say overall it's a very respectable set of results. All our four associates, despite the initial concerns that we have with inflation and weakening currencies, I think across the board we have seen actually very positive mobile revenue growth. and profitable growth as well. So all four of our associates have fired up, and uh and I think all that added up to a very decent performance in the first half.
5:00Adam: Alright, it's good to hear. So I mean I think shareholders will be happy to hear that.
5:03Unidentified speaker (speaker_2): I hope so. Yeah,
5:05Adam: I think so.
5:06Unidentified speaker (speaker_1): So I I mean I can I can feel your confidence, you know, when
5:09Unidentified speaker (UNKNOWN): you when
5:09Unidentified speaker (speaker_1): you were talking about the guidance because I uh I went to a lot of Azjax events and
5:14Unidentified speaker (UNKNOWN): Azjax Ash
5:14Unidentified speaker (speaker_1): Jax people. always tell me, you
5:15Unidentified speaker (UNKNOWN): know,
5:15Unidentified speaker (speaker_1): uh actually the company can give guidance on, but a lot of people don't dare give. But I see you
5:19Unidentified speaker (UNKNOWN): can even
5:19Unidentified speaker (speaker_1): upgrade your guidance to
5:20Unidentified speaker (speaker_2): correct, right? It's it's uh of course it's always a risk, right? Because you don't know what will happen, but I think we do feel confident. Yeah, and and I remember uh two years ago you always said that it's important to tell the investors, right? Guide them positive, but also if things are difficult, let them know it's difficult, right? And be open to them.
5:37Unidentified speaker (UNKNOWN): Yeah,
5:37Unidentified speaker (speaker_1): so I remember I attended your I think one of the analyst briefing,
5:42Unidentified speaker (UNKNOWN): I think
5:42Unidentified speaker (speaker_1): I think the first quarter or something like that. I can't remember when.
5:44Unidentified speaker (UNKNOWN): Recent, this year.
5:45Unidentified speaker (speaker_1): Uh this
5:45Unidentified speaker (UNKNOWN): year, yes, right. Okay.
5:46Unidentified speaker (speaker_1): And actually, you were saying that the optus was was I think the optus in charge was saying that it's
5:53Unidentified speaker (UNKNOWN): tough,
5:53Unidentified speaker (speaker_1): inflation very high, everything was
5:55Unidentified speaker (UNKNOWN): very tough.
5:55Unidentified speaker (speaker_1): And yet there's a positive EB e-bit growth over here.
5:58Unidentified speaker (speaker_2): Yes.
5:58Unidentified speaker (UNKNOWN): Yeah.
5:59Unidentified speaker (speaker_1): So
5:59Unidentified speaker (UNKNOWN): that's why I was
5:59Unidentified speaker (speaker_1): that's why Jessica Adam Edd were asking me, you
6:01Unidentified speaker (UNKNOWN): know,
6:01Unidentified speaker (speaker_1): hey, correct, how's the result? Is Yeah. it?
6:02Unidentified speaker (speaker_2): Yes.
6:03Unidentified speaker (speaker_1): Good.
6:03Unidentified speaker (speaker_2): Good. Yeah.
6:05Unidentified speaker (UNKNOWN): In fact, you're reading that. But you're
6:06Unidentified speaker (speaker_2): right. The Australian business uh it continues to be challenging, right? Whether it's inflation, whether it's just competition and all that. And and to be frank, we had our own issues, right? And and we are definitely not out of the woods. We continue, we cannot take things for granted. We have to make sure our network is solid, you know, it's functioning well, our customers that we in initially, I would say in the last few years we might have lost a lot of customer trust. Uh, we need to earn that back, right? But we're on the right track.
6:34Unidentified speaker (speaker_1): I see you are managing the cost well because the revenue is dropped 2%, but the EB actually went up, so
6:38Unidentified speaker (UNKNOWN): you're imagining the
6:39Unidentified speaker (speaker_1): good cost. Correct.
6:40Unidentified speaker (UNKNOWN): And
6:40Unidentified speaker (speaker_2): the cost is because you see, there's uh revenues you got like our mobile services revenue. Some part of that revenue is also selling uh handphones, right? Uh uh mobile phones. So that kind of equipment plan is very dependent on whether we launch the Apple phone or Samsung phone, right? So that's quite lumpy. So first half of the year don't have new phones. Only Apple 18 just came out, iPhone 18 just came out a few weeks ago. So you'll see that being recognized in the second half of revenues. That's why revenue comes down a bit. But the fundamental EBITDA growth is there, and the mobile service growth, which is taking out the equipment plan, has actually grown.
7:15Unidentified speaker (speaker_1): Yes. Good to hear that for an
7:17Unidentified speaker (UNKNOWN): author. So
7:17Adam: there's a whole lot of uh stuff that you just shared. Yes. The one thing I want to ask you right now is that, like Victor was mentioning, not many companies give guidance, but you're one of the few that do. And you're guiding that uh opco e-bit is gonna be single to high uh single
7:33Unidentified speaker (UNKNOWN): digit. Single digit growth.
7:35Adam: How do you plan to achieve that?
7:37Unidentified speaker (speaker_2): Okay, we plan to achieve it, I think. let's if you talk about the opco growth, right? The four opcos that we have is SingTel Singapore, Optus, Digital Infractor, and NCS. So let me start with STSG. STSG is continue to compete in this very challenging market, focusing on fundamentally the customers that we serve across the three brands. So early this calendar year, or maybe late last year, SingTel actually came out with what we call the tri-brand strategy, three different brands, targeting three different customer base. So one, we have the full service brand, right? And that's branded SingTel. And that one we we uh the price point is is the highest among all the three. Uh and but we we we give you know 5G plus priority service, we added a lot of value-added services. mobile security, and everything, and that's how we targeted it. Then there's a second brand called Gomo, right? Uh and this Gomo brand is digital only, targeted at the youths who are more digitally savvy, and also uh digital seniors who are very comfortable with just doing digital, right? And seniors sometimes can be a bit more price-sensitive, price sensitive, right, because they want a cheaper plan. So Gomo. is priced somewhere in the middle. And then we've got the high plan or the
8:56high brand, which is the value conscious buyer. The value conscious buyer could be again seniors, could be students, it could also be uh migrant workers, right, who are value conscious, understandably so. And every segment, right, we cater to the needs of the customer. And every segment we have a different uh price plan, we have a different offering, and really targeting to know what the customers are. And what they do is all three uh kind of go out there and really do what's best for that particular customer segment. And we have actually seen that to be very, very effective. All three brands, you know, customers have as we on the consumer side, we've actually seen uh additional customers during this period. Uh or what we call net ads, right? We have included that uh and and we've seen revenue growth, as I've said, right? So that's what we are doing. The other bit is also enterprise. Don't forget, uh, SingTel is not just on the consumer business. More than half of our revenues actually come from the enterprise business. So this is taking care of big enterprises as well as small and small and medium enterprise, not just providing mobile uh services. but we also provide connectivity. For example, we own a lot of the fiber underground. We own a lot of submarine cables. So a multinational company like Nestle needs a lot global connectivity and we actually own this global
10:16kind of portfolio, right? Or network of sub-sea cable. So we are seeing a lot of interest there, especially in the data and internet space. Data centers, you know, be we you you you've got you got two data centers, they need to be connected, and not just a Singapore data center as we're connected to the Thailand data center. It's all through terrestrial cable or terrestrial fiber or sub-sea cables, right? And and we own that. So that we're actually seeing an improvement in the business there. Opters, I mentioned, just keep focusing on the basics, making sure our network is secure, our network it continues to be high performance. performing, and also we win back the customer trust, keep focusing on what's best for the customer, and don't give them the impression that we are, you know, just going after profits only, but not taking care of the customer. And then from that, you know, and and recently in May, they upped their prices a bit. So we're starting to see that flow. So I see this flowing into the second half. Our enterprise business in Australia is also starting to do quite well. You might recall two years ago when we were
11:16sitting down, it was actually loss making. Today it's turnaround and actually gives us some profit.
11:21Unidentified speaker (UNKNOWN): Then
11:21Unidentified speaker (speaker_2): DICO, which is our basically our data center business, Nexera, as well as REAI, which is our GPU as a service. This is an area where we need to focus on growing, but we need to grow it in a very measured way, especially as what we're seeing, all the volatility. that we're seeing in the market today, right? We need to focus on number one, high quality customer contracts. If we don't have a customer contract, we don't build the data center or we don't add in the equipment in the data center. Uh GPUs as well. We don't add on GPUs and then just hope, oh, we'll get customers two years from now. If there's a customer contract, secure, high quality, hyperscaler type of customers, we will do the business. The other thing that's very important today, especially for data center business, we need powered land. Meaning the land is not only just land, but we need to make sure we secure power and we secure water, right? Especially for AI DCs. So that's very important to us. We want to grow that, but it cannot be grown at all costs. We need to make sure we have a very disciplined approach to CapEx because especially for data centers, the money just keeps going, you can if you're not careful, you just within a year spend like two, three billion dollars, right? So that's something we are very focused on. And we're seeing the growth coming because uh we opened up DC TWAS early this year, right? Right now, I think we're
12:41probably uh we're and this kind of thing we need to fill floor by floor. DC TWAS, I think today, uh as of our first half, it's probably we're at 30-40 30, 40% percent fill up. I think the second half we're starting to see go to 80%, and then maybe another three, four months, 100%, right? It's all 100% filled already, but it just takes time for customers to move it. So you'll actually see a very nice ramp up in income because DC TWAS is quite big. REAI, we have GPU as a service business. We want to focus right now only in Singapore because it's high quality Uh and our we contract a lot. with, I would say, uh a lot of uh different government type agencies and large companies. And we're s we've we said early this year we're deploying uh 300 million for six megawatts. Uh we've deployed four of the six megawatts already, and the four is already generating income as of uh one, two months ago. So we're starting to see that wrap-up. wrap up.
13:36Unidentified speaker (speaker_1): Yeah.
13:36Unidentified speaker (speaker_2): Yeah.
13:36Unidentified speaker (speaker_1): So I
13:37Unidentified speaker (UNKNOWN): think
13:37Unidentified speaker (speaker_1): maybe you can I think the Nixera and the REAI is the most interesting because they
13:42Unidentified speaker (UNKNOWN): are the
13:42Unidentified speaker (speaker_1): old I think the only one or something that that gives twenty twenty-while percent. EBIT
13:47Unidentified speaker (UNKNOWN): growth, right?
13:47Unidentified speaker (speaker_1): EBIT growth, okay. Yeah. So maybe you can give our viewers sorry, uh, what's the exact difference between the Nixera and REAI? Can.
13:56Unidentified speaker (speaker_2): Okay. So uh let's start with data center, right? Nexera. Next era. Think of it as like uh a building, right? So we actually put up a building, right? We build a data center. So data center is not just a standard building. You need to make sure the floor is specially there's reinforced concrete, and then there's ability to connect to fiber and all that because it's you are going to put uh what we call Called racks on it, right? And compute uh power. You need to also organize the building such that you can secure different parts of different floors. Because certain customers, especially when it's sensitive workloads, they might be very uh concerned that if if uh let's say the data center is six floors, that if anybody walking into the third floor can easily go to fourth floor, go to fifth floor, right? You need to have a secure environment, right? And also from a connectivity is secure. That means you are not linked to an external network which can be easily hacked, right? Which or potentially can hack, not easily hacked, but potentially can hack.
14:57You've got to keep it within the ring fence, you got to contain it, right? So that you don't have intruders, cyber intruders coming in. Then at this, then what you do is when we offer the space, so it's not just space, but then we offer, we build racks, Recks and then they start bringing in their machines, their compute, and then that's where the customers working with us. as the landlord in some ways actually then they pay us a rental fee. So usually the data center leases can be five years, but most of the time it's seven to ten years. So very long data type. And uh and and there's uh escalation in terms of every year, there's some inflationary cost uh adjustment. Uh so it's is uh so we are inflationary protected. We are also power price protected, it means there's a power. pass-through. So if our energy costs go up, the customer will pay for it. So that's generally the model of data centers. GPU as a service is part is is an addition add-on to the data center business. So we talked about building racks for our customers, making sure it's secure and everything. Then to actually have the machines, you need chips. Okay? And the chips will then be used to process whether it's a cloud compute or whether it's AI compute, right?
16:07And the workloads typically, some customers, like the hyperscalers, may say they do their own compute. But there are a lot of customers other than the hyperscalers. And when I talk about hyperscalers, we're talking about the Amazons, the Microsoft's, Microsoft, the Googles of the world, right? But there are only five or six of them, right? And then maybe throw in some of the Chinese. very most, 10, right? Very most. But you've got so many enterprises, the banks, you know, agencies, universities who actually need to do AI or they need cloud storage, right? So, but they say, look, I don't want to be buying the chips. I'm not sure how much I need two years from now. Everybody is going to AI, but do I need X capacity or do I need Y capacity? So what they do is they work with us. They say we tell them, don't worry. Our REAI team will go in there, don't worry. We will take care of all the AI infrastructure and the services for you, right? You but you must pay us a fee, right, for that. So that's what we do. So we procure the chips from let's say an Nvidia, we install the chips, we run the workloads on behalf of our customer, they pay us a fee for providing that service. So, this is what you mean GPU as a service. That means we take on the GPU chips, right? But we rent out and release out. So typically the contracts for GPUs is about three to five years, right? A bit shorter than data
17:27center. But the good thing is when we pay for the chips, let's say today, day one, within six to nine months, we can get revenues ready because it's very fast. We get the chips from let's say Nvidia, we install it. takes about six to nine months, we can start generating revenue. Data center is a bit different. Data center where we spend, where we break ground, we build a building. It's only three, four years before custom when it's building finished ready, like a construction site finished, then we start generating revenue. So the cash flows are quite different. The margins, it depends, but EBIT margin is in the 20s to high 20s, twenty uh low 20s twenties to high 20s twenties for both businesses. but the time to cash is much faster for GPUs than for data centers.
18:08Unidentified speaker (UNKNOWN): All
18:08Adam: right. So the two of them are actually very complementary businesses.
18:10Unidentified speaker (UNKNOWN): It's complementary, but they're
18:12Adam: different. So
18:12Unidentified speaker (UNKNOWN): one of
18:13Adam: them is like a landlord, like Nexaro is like a landlord. In some
18:16Unidentified speaker (UNKNOWN): data
18:16Adam: centers in that sense. Correct. Then REA
18:18Unidentified speaker (UNKNOWN): is
18:18Adam: almost like a turnkey solution.
18:20Unidentified speaker (UNKNOWN): Like a
18:20Unidentified speaker (speaker_2): turnkey service provider. Correct.
18:21Adam: to go
18:22Unidentified speaker (UNKNOWN): to the project. And they work together
18:22Adam: to That's lots a lot of synergy in that sense.
18:25Unidentified speaker (speaker_2): Although the businesses are kept separate. deliberately. So KKR, for example, is a minority owner in our Nexera business, but REI is 100% owned by us.
18:35Adam: So you previously mentioned plans to monetize Nexera.
18:39Unidentified speaker (speaker_2): That's right.
18:40Adam: How's that coming along? Is that
18:41Unidentified speaker (UNKNOWN): uh
18:41Unidentified speaker (speaker_2): yeah, that while the it's ongoing, right? We had uh actually Actually, have we been having discussions uh in the past few weeks. Um uh okay, monetizing Nexera. can come in many different ways, right? Uh it could be the most straightforward is just do an IPO, right? And the IPO can be in Singapore. We are exploring other options outside Singapore because uh it will be quite a large IPO. Singapore sometimes have constraints on the size of the IPO. So we are exploring a US IPO and maybe other jurisdictions as well. But that's the most straightforward option, right? We could uh we We could reorganize this, reorganize the businesses. We could have a business that has stabilized assets, and the other business is more development assets, right? So we are exploring different ways. We're still very early stage. But also don't forget, right? Other than Nexera, today we have another platform called STT GDC, which is much bigger in size than Nexera. Because the owners of both of these. platforms are similar, we have actually a lot of flexibility in what we want to do with both platforms. So we are still working out. We just acquired or completed the GDC transaction about slightly more than a month ago. So we're starting to think what we can do. But it's amazing because the combined capacity, right? Just focus on
20:01the current operating capacity. combined is 900 megawatts, almost one gigawatt of capacity, right? And not only that, what's also included important is the contracted capacity that's coming online in the next few years that's already contracted by customers. And that's over one over one gigawatt. I think it's like 1.5, 1.6 gigawatt of contracted capacity. So these are customer contracts in place already, we're just building it. out. So it's a very exciting platform and we're very excited to see what we can do here. Yeah.
20:32Adam: All right. So I mean these are massive investments in
20:35Unidentified speaker (UNKNOWN): greater
20:35Adam: centers. And you mentioned just now, like it's it's a lot of capex. Yes. Um is it a little bit risk? How what how do you deal with the risk at this point in time? Because
20:44Unidentified speaker (UNKNOWN): everyone's talking about
20:45Adam: the AI bubble.
20:45Unidentified speaker (UNKNOWN): Yes,
20:46Unidentified speaker (speaker_2): uh I and we are watching that very carefully, right? I think fundamentally we do believe the demand is still here to stay. Because you know what, demand now for data centers, for all these compute is not just driven by the five, six big AI players, right? I mean, don't forget there's still so many companies are starting to discover how to use AI, right? How to embed AI into their businesses. And this is not just providing employees with Chat GPT, right? It is more than that. It is embedding into the business. How do you make how do you upsell the customers? How do you run customer service? How do you do procurement? How do you run your IT services? For us network, how do you use AI to start predicting congestion in the network, which is extremely powerful? So all this would require data centers, REAI, chips, and all that. We are focused on that kind of business. We are not focused. on really, you know, just servicing all the big AI plans, you know, the what you read about in the in the US. So the approach we are taking is number one, high-quality customers. Very, very important. When we want to put comic care packs, we always want to make sure we've got high-quality customer contracts that's contracted. Number two, if it's a data center we're going in, in this data. day and age, power and water is
22:05something that we need to make sure is secure. And more importantly, we engage with the community in the local community to make sure this is something that is additive to the community rather than something that will, you know, take power from them or take water resources from them. Very, very important. This one we take very seriously. SingTel stands for a much higher standard of what we want. to do with the community and with society. Finally, it's also capital structure. Right? You hear of all this circular financing, this financing this, who in turn finances back. No, no. Ours is very simple. We will finance with bank support, with maybe bonding, a bond support, and our own balance sheet, SingTel balance sheet. Very straightforward, no circular financing or all that kind of thing. We will finance it in our balance sheet. Now, the flip side is you hear all these big deals getting done, oh, I'm doing 100 megawatts. We are doing it in a very measured way and deliberately sold up. Even in that measured way, we are confident of growing the business over the years, but it's not those crazy types of growth. So it's a very measured type of growth, very disciplined approach to capex. Yes.
23:11Adam: Alright, on the theme of risk in and competition, now this is one question I really want to ask. you. Okay. I'm highly interested in Starlink.
23:18Unidentified speaker (UNKNOWN): Ah,
23:18Unidentified speaker (speaker_2): yes.
23:18Adam: Yes. So uh recently Starlink announced uh plans to be basically become like a telco.
23:25Unidentified speaker (UNKNOWN): Yes.
23:25Adam: Right? Now the advantage of Starlink is that they are global. Yes. And
23:30Unidentified speaker (UNKNOWN): people
23:30Adam: watching this, they may kind of think that oh Syntel is uh you're I mean the mobile business is still a key part of your business. Yes. Uh but some people may not. remember that or understand that SingTel is not just Singapore. You have you have a mobile business in in uh India, Indonesia, these are huge countries. So in Singapore, it's just like one city, easy to cover. But in these like major like geographies, could Starlink come in and actually disrupt your telco business?
24:00Unidentified speaker (speaker_2): Very good question. It's something that we are talking about it almost every other every week, right? Talking about strategically how we should look at satellite technology. Um and this satellite technology is different from broadcast satellites. This is really communication, right? So LEOs they
24:17Unidentified speaker (UNKNOWN): call
24:17Unidentified speaker (speaker_2): it, right? Low Earth orbitals. Or Starlink is definitely one of the companies that are involved in satellites that that we have to watch. Now, satellite technology is very much part and parcel of communications, mobile communications especially. Now, that satellite technology, you know, if you see it, you have uh a mobile player like us. We have to see, okay, is it additive, is it complementary to what we provide, or is it a competition? Now everybody thinks it will be competition. Maybe in the US it's positioned that way. For us, it's a bit more nuanced, a bit more um not that straightforward. And I'll explain why. If you look at uh let let's look at uh satellite broadband and what we call D to D, that means direct to device, right? Let's look at broadband. If you look at satellite broadband today, it's actually good for telcos to partner with satellite companies. Because for a very big country, let's say geography, Australia. right? Australia today is probably 95% of the population is in 5% of the land mass. The other 95 is you know the big
25:25Unidentified speaker (UNKNOWN): unknown, right? Yes,
25:26Unidentified speaker (speaker_2): the outback, right, as they call it. So using just pure mobile towers to connect is It's commercially, it it's just it's just not the right thing to provide to the customers. With satellite communication technology going into the rural and the more remote areas, it's definitely a big plus to partner with a satellite company to provide that. Right? So from a broad and and and that's on the kind of mobile side, but also on the broadband side. You know, we have fiber going what we call fiber to the home. But in remote areas, it's also non-economical to lay fiber out into the remote area just to cater to one house. right? So so using satellite technology there is actually makes a lot of sense. And we would encourage and we'll definitely explore any form of partnership with a satellite company. Now, but you know, at the same time, I think the other thing we need to be mindful in more urban areas, the physics of satellite may not be suitable to a very dense area, let's say like Singapore, or even uh or even you know. many cities in India, right? Because it's very dense, very uh uh you know the it's an urban population. So not only you have population density, it's very hard for satellites to to adequately cover a very dense area. And also, equally important if it's a build-up area, whether it's
26:46MLT trains and underground or whether in a building like this, uh satellites. satellite communication technology or connectivity may not be as smooth or as well covered as mobile technology. It's just pure physics. Now, things may change, there could be technology, but you know, I think it's probably in the next 10 years, perhaps after that, then you maybe technology may change. The other thing that we need to be mindful about is also regulation. You see, yeah, when we when a satellite company comes in and wants to provide uh mobile communications, they could use two types of things, what we call spectrum. It's basically space or air, right? It's a air space, right? To you know, frequencies and all that to able to conduct uh communications. They could take the spectrum that mobile companies every you know every now and then there's an auction, we own the spectrum, they could easily go. there and participate in auctions uh and and buy over the spectrum. Or they could use satellite spectrum, which is available as well. Right? So that is possible, they can use the spectrum. So so but some countries uh uh the governments, the regulators are a little bit more careful about having foreign satellite companies to own the spectrum because it is actually very critical. There are some countries. who said only local operators or
28:05locally majority owned companies can own that spectrum. I mean you have heard reports where in certain cases a global satellite company will just switch off and all communications are disabled. So I think from a government regulatory perspective they would have to think through these issues uh because it is not straightforward. right? Uh uh and and in addition to spectrum, there's a whole slew of different regulations that need to be thought out in order for an international or foreign satellite company to provide connectivity services uh uh across the whole population. So that one also, you know, spectrum ownership, regulatory approvals, you you really have to go all through all that, right? Uh in the US, I think you have heard uh Starling or SpaceX uh looking to buy spectrum in the US. It's an American company looking in the US, it's definitely more straightforward than you know an American company looking at a foreign country, right? Because then it gets a lot more sensitive. Um and then finally, I think there's a commercial argument.
29:07Unidentified speaker (UNKNOWN): You
29:07Unidentified speaker (speaker_2): see, um, in in mobile business, when we look at communication, we always look at APU, RPU, which is the revenue per user.
29:14Unidentified speaker (UNKNOWN): In the U.
29:15Unidentified speaker (speaker_2): US, an average RPU for all the telcos there is $50, $60 US. The cost per user for satellite is probably less than that. So if it's let's say, example, $30. If your RPU is $50, your cost of providing satellite is $30. The disruption is the $20.
29:35Unidentified speaker (UNKNOWN): The
29:35Unidentified speaker (speaker_2): RPUs in places that we operate operated, in, India. Thailand, or India, Indonesia first, Philippines. Make a guess how much is it? ARPU.
29:44Unidentified speaker (speaker_3): Must be even lower.
29:46Unidentified speaker (speaker_2): $3 US. Roughly US dollars.
29:50Unidentified speaker (UNKNOWN): So
29:50Unidentified speaker (speaker_2): can satellite technology have a cost structure that is below $3? Because if the cost is $20, then it's just non-economical to go. Of course, there are some countries like Singapore where the APU is higher. But then again, we are very dense population. Is there can a satellite communication technology be able to give as good, if not better, service, connectivity service than a mobile provider, a terrestrial mobile
30:14Adam: provider?
30:15Unidentified speaker (speaker_2): So it's really depending on the markets, depending on not just the geography of the market, you know, in terms of the land mass and how big it is, but its population density and also the ability for the user. in the country to pay for it. So it's a complicated question, uh, but I it's also got a complicated answer. But look, we're not saying that we you know, it's not a threat, right? Satellite uh technology, we are watching it, but personally I don't think it it needs to be a competitor. We can be a partner, right? Uh and it helps our business as well and our customers. All
30:49Adam: Alright, right. So it's been really challenging in the mobile market in Singapore for a few years. But I think things are maybe looking a little bit better better because there's consolidation in the industry. Looking in direction.
31:01Unidentified speaker (speaker_2): We hope so, yes.
31:02Unidentified speaker (UNKNOWN): What
31:02Adam: is your view moving forward about that? And do you have does like does SingTel have any plans to acquire any
31:08Unidentified speaker (UNKNOWN): Okay.
31:08Unidentified speaker (speaker_2): Should I wear my hat of uh a SingTel CFO? Or should I wear the hat of a customer? Maybe I'll wear the hat of customer first.
31:15Adam: Okay.
31:15Unidentified speaker (speaker_2): Of course, customers will want as cheap
31:17Unidentified speaker (UNKNOWN): data as possible, right?
31:18Unidentified speaker (speaker_2): I also want, right? Uh when I go roaming, I want free roaming, whatever you can give me free, I will just take, right? Um, but now I change my hat. I wear a uh uh let's say a singTel CFO. You see, if you look at the day and age we live in today, if mobile prices continue to go down and data prices in Singapore are the lowest in the world.
31:39Unidentified speaker (UNKNOWN): In
31:39Unidentified speaker (speaker_2): the world, right? We can show you the statistics given all the price competition. Now, in the short term, it's of course very good for customers. My mother, my father, my wife and all my family are all very happy, right? But the issue is over time, can the operators earn a decent return to be able to make their the business sustainable? Again, I asked, look at our competitors, how much profits they are earning. I can tell you personally, I don't think it's sustainable. And why is that important for the long term benefit of Singapore? It's because if you look at mobile business today, telco, connectivity business, it has become what we call critical infrastructure. Right? Think of it in terms of what you're you doing. during COVID years, right? How important connectivity is important? So if the operator is losing money, how can you expect the operator to reinvest into the network and strengthen the network? And you strengthen the network across various things. Number one is cybersecurity. Is the operator spending enough to protect the network, especially with in the world of mythos today,
32:49where you can just send a hundred agents. just to create havoc in the network. That is serious stuff, right? And in order to protect the network, you've got to spend money, like it or not, right? So number one, cyber cybersecurity, right? Number two is AI. If we want the network to be AI smart, as I said, predict where the congestion is, predict what we can do better for our customers. We also need to spend money. Operators also need to spend money. But over time, I'm afraid with too much competition, if operators lose money, the decision will be: I want to maintain my profit, I do not want to invest. Right? And then they take a very short-term view. And in the long term, for Singapore, it's actually a negative because one day, if we get attacked, it It can be quite bad. So I do hope we will reach some middle point. I'm not suggesting that they. data prices go up. As a customer, I hate I is it's not good, but also it's not right for our customers, right? But it has to be reached a stage where it's sustainable for the industry. And that's something we need to focus on.
33:50Adam: Alright, so that's
33:51Unidentified speaker (UNKNOWN): good
33:51Adam: to know. So I
33:52Unidentified speaker (UNKNOWN): I
33:52Adam: think we've covered quite a whole bit of your results. Yes. Uh you're doing tremendously well. Uh congratulations on your
33:58Unidentified speaker (UNKNOWN): oh thank you.
33:59Unidentified speaker (speaker_2): Thank you very much. It's it's a great teamwork by everyone.
34:01Unidentified speaker (UNKNOWN): Yes.
34:02Unidentified speaker (speaker_2): And
34:02Adam: but we also want wanted to cover a bit of the competition and the risk, so like Starlink, uh you know the AI bubble, and of course competition in Singapore in the mobile yes um area as well. Um
34:12Unidentified speaker (UNKNOWN): a
34:13Adam: lot of investors will be looking at SingTel because of the dividends that you pay out.
34:16Unidentified speaker (speaker_2): Correct.
34:17Adam: So um like I said like I mentioned earlier, not many companies give guidance on dividends, but you're one of the few that do.
34:22Unidentified speaker (speaker_2): We have decided to do that, yes. Yeah.
34:25Adam: So I think that that's a good thing. I think I think
34:28Unidentified speaker (UNKNOWN): leading
34:28Unidentified speaker (speaker_1): by example,
34:29Unidentified speaker (UNKNOWN): what
34:29Unidentified speaker (speaker_1): SGX wants companies to do.
34:31Unidentified speaker (UNKNOWN): Yeah, yeah, correct,
34:32Unidentified speaker (speaker_2): correct. And and we hope that um we hope that through this that we encourage other companies to do it. But more importantly, also from an investor standpoint, whether it's institutional investor or retail, they will then you know spend the time to really focus on can the company sustain the dividend, can the company make profits, you know. things like that. Because you you cannot just listen to your buddy and say, hey, buy this stock or buy that stock, right? We have to do the work. Uh I think you might recall two years back when we talked about we introduced this concept of a VRD, value realization dividend. And I I happy to say we were the first in the market, and then after that there were a few companies, they call it differently, but effectively it's a VRD, right? Uh we hope to be also uh uh one of the first companies to give guidance on dividends. So very unforeseen circumstances, we are looking to grow our full-year dividend by mid to high single digits. So you translate that to mid to high, call it 19 and a half to 20 cents, like that, right? You just from last year. Because last year we paid 18 and a half cents. So that's the guidance we are giving, right? And And we hope that would help steer and show. the investors you know our commitment and also our our signaling the outlook that we're seeing. The next six months for this financial year will continue to be difficult. Our FX continues to I mean
35:52Sing dollar continues to be very strong. Competition will continue, I don't know when you know there will be consolidation in Singapore, Uh the AI, you know, we have to navigate very carefully and very disciplined in our cap packs. But we continue but we believe we have a good team, we have a right strategy to continue to execute. And and if we do well, we as we've always said, we want our investors to come along and share the benefits with us la. So so that's why we are giving that guidance.
36:19Unidentified speaker (UNKNOWN): Yes.
36:20Adam: I think a lot of uh investors will be very happy to hear that.
36:22Unidentified speaker (speaker_2): I hope so, yes.
36:23Unidentified speaker (UNKNOWN): Yeah. I think and
36:23Adam: it it's sets a good example as well for the Singapore market. Yes.
36:27Unidentified speaker (speaker_2): I tell you, we debated this so much. Like, truly, what can we say? And I think we feel quite uh quite confident we can deliver, barring unforeseen.
36:36Unidentified speaker (UNKNOWN): I mean,
36:36Adam: unless a COVID situation or something like that happens. But it's good to know that you're giving that confidence to the investor.
36:42Unidentified speaker (speaker_1): Yes.
36:43Adam: And that gives us confidence to like, okay, I I believe in the management, I
36:46Unidentified speaker (UNKNOWN): believe
36:47Adam: in the company, Yes. and I can have some sort of you know
36:49Unidentified speaker (speaker_3): predictability.
36:50Unidentified speaker (UNKNOWN): Yeah,
36:50Adam: predictability.
36:51Unidentified speaker (UNKNOWN): Yeah, that's important.
36:52Unidentified speaker (speaker_2): Predictability. Especially in this world of unpredictability. I mean, you know, things are really like a roller coaster on a daily basis, not even a weekly basis, right? Um and and to your point, I think the key is really mean what you say and say what you need. Right? Um early this year when we announced our full year results and we gave guidance for the year, I mean a lot of investors were a bit shocked. Said, Wow, why do we give a guidance of low to mid? But at that time, you must understand we we really felt that we had to call out the risk because Homoose just happened a few weeks ago, right? Before that announcement. And then you had the Simba and M1 merger calling called off, which meant there's going to be ultra competition in the market. And and uh and because of all this, we thought that you know, better to understand. underpromise and then over deliver like yes correct
37:41Unidentified speaker (UNKNOWN): yeah
37:42Adam: all right I think that's a really good um you know update of what's been happening
37:47Unidentified speaker (UNKNOWN): yes yeah
37:47Adam: uh thank you so much for having
37:49Unidentified speaker (UNKNOWN): us to
37:49Unidentified speaker (speaker_2): see you guys again
37:50Unidentified speaker (UNKNOWN): answering all our
37:51Adam: questions
37:52Unidentified speaker (speaker_2): but you know we should do this more often yeah
37:54Unidentified speaker (UNKNOWN): yeah and it's
37:55Adam: I think it's a good time for the updates. Two years is a good time to see how the business has been doing. You have some time to execute all your plans. True
38:01Unidentified speaker (speaker_2): true
38:01Unidentified speaker (UNKNOWN): and the
38:01Adam: results have have shown shown
38:03Unidentified speaker (UNKNOWN): yourself. Thank you very much. Thank you so much. So you say
38:07Unidentified speaker (speaker_1): you do what
38:07Unidentified speaker (UNKNOWN): we're doing. So thank you
38:08Adam: so much, Arthur, for being uh here today. And thank you for having us
38:11Unidentified speaker (UNKNOWN): at your studio.
38:12Unidentified speaker (speaker_2): Thank you. Thanks for catching up. So
38:13Adam: with that, uh thank you so much for watching. If you have any questions or comments, put them in the comment section. We're happy to answer them. Um of course if you like this round table, please hit the like button and subscribe to our channel. Many more round tables coming up, and we'll see you again.