Title

Presentation

Hee Jun Chung: Good afternoon. I am Hee Jun Chung, IRO of SK Telecom. Thank you for joining SK Telecom’s earnings conference call. Today, we will first deliver a presentation on major events, 2024 Q3 earnings highlights and business performances, which will be followed by the Q&A session. The call is attended by our executives from relevant business divisions, including Yang-Seob Kim, CFO of SK Telecom, and we are providing consecutive interpretation for the call. I would like to remind you that all forward-looking statements are subject to change, depending on various factors, such as market and macro situations. Let me now present our CFO.

Yang Seob Kim: Good afternoon. This is Yang-Seob Kim, CFO of SK Telecom. After the spin-off in November 2021, SK Telecom announced a new vision to focus solely on telecommunications and AI. For the past 3 years, the company has established an AI Pyramid Strategy, mobilized enterprise-wide capabilities to become a global AI company and pursued a number of partnerships and investments. As the recently announced Corporate Value-up Plan illustrates, SK Telecom will now show you specific ways to make money using AI. There are three main ways to generate profit with AI that SK Telecom is pursuing. First is AI data centers. In the D.C. market, demand triggered by AI continues to exceed supply. We have expensive DC operational experience, AI solutions secured together with domestic and global partners and nationwide base stations that can be utilized as AI infrastructure.

SK Telecom aims to become a leader in the global AI data center market. Second is AI B2B. Enterprise business is expected to see rapid market expansion in the AI era. We will actively utilize generative AI foundational models secured through in-house effort and external partnerships and mobilize B2B capabilities available within the group to identify brand use cases and win contracts. Third is AI B2C. The B2C service market has been led by platform companies based on advertisement business models. However, advertising revenue is not sufficient to cover the cost of building infrastructure needed to provide AI services that are becoming popular recently. As such, subscription business models are gaining attention. Given this trend, new opportunities will be opening up for telcos that already have subscription business models.

SK Telecom has already established a strong subscription model with T Universe, and we ask for your support and interest in how we will advance our A. and global personal AI agent services. Let me now report on the consolidated financial results for Q3 2024. Consolidated revenue reported KRW 4,532.1 billion, up 2.9% year-over-year, driven by the stable growth of roaming and B2B solution business. Operating income posted KRW 533.3 billion, up 7.1% year-over-year, thanks to the stabilization of costs such as marketing and depreciation. The positive results show the impact of strengthening business fundamentals through operational improvement and enhancing efficiency with a company-wide adoption of AI. Net income posted KRW 280.2 billion. On a nonconsolidated basis, revenue posted KRW 3,203.2 billion, up 1.7% Y-o-Y.

Operating income and net income reported KRW 457 billion and KRW 222.4 billion, respectively. Let me now turn to business highlights for Q3 2024 and strategic directions. First, data center revenue grew 14% year-over-year, thanks to higher utilization of data centers. In terms of AI monetization, AI DC is the business that can expect the most short-term results. We aim to be a leader in both the Korean and global DC market by applying emerging cooling technology of SK Group for energy efficiency, supplying GPU power of Lambda effectively and employing AI data center solutions of Penguin solutions. As we announced at the AI Summit this week, we will establish the AI infrastructure super highway to evolve it as core infrastructure that will accelerate the development of Korea’s AI ecosystem.

The super highway will enable nationwide connection through Edge AI based on AI data centers as regional hubs and GPU as a Service in the Seoul metropolitan region. We are currently developing the telco-specific LLM together with the members of the Global Telco AI Alliance. We aim to launch it commercially in the first half of next year, and we’re in the process of establishing a global AI joint venture for the LLM service. We plan to first apply the telco-specific LLM to our own customer services through customer contact centers and T world to assess its performance in effect. Next, regarding AI services. We implemented major changes to A. in August to provide natural conversational experience and professional services through diverse agents and strengthen search and other everyday convenience features.

Thanks to the overhaul, the cumulative downloads as of the end of September exceeded 5.5 million. In October, we upgraded T Phone to A. phone with AI phone features and launched the PC version of multi-LLM agent. We plan to expand the application of A.’s AI features to various services. In the MNO business, our 5G subscribers recorded 16.58 million as of the end of September 2024, accounting for 73% of the total subscriber base. SK Broadband secured 9.62 million pay TV subscribers and 7.11 million broadband subscribers as of the end of Q3. Our strategy for the fixed and mobile business aims to improve profitability through a stable market operation while defending the top line by focusing on attracting and retaining high ARPU subscribers. In Q3, enterprise business sustained a growth trend of 8% year-over-year.

The cloud business also saw a revenue growth of 30% year-over-year with stable fixed line business performance. Enterprise AI business continues to win meaningful orders, and we will grow the enterprise AI business as a key pillar of B2B business and develop concrete ways to generate profit. Finally, let me turn to shareholder returns. The DPS for Q3 was determined at 831, the same as Q2. For the remainder of 2024 and beyond, SK Telecom will strive to maximize corporate value as well as shareholder value. We look forward to your continued support and encouragement. Thank you. We will now begin the Q&A session.

Operator: [Foreign Language] [Operator Instructions] [Foreign Language] The first question will be presented by Seung Woong Lee from Yuanta Securities.

Seung Woong Lee: I’m Lee Seung Woong from Yuanta Securities. I would like to ask you a couple of questions. First of all, when we look at the earnings results for the third quarter, we see that there’s a larger recognition of nonoperating losses, which has led to a smaller net income. I wonder if this will have any substantial impact on your year-end dividend? The second question has to do with your corporate value-up plan, which you announced recently. However, SK Telecom is not included in the value-up index. So I would like to understand your responses to that situation and the background behind your disclosure of the corporate value-up plan?

Yang Seob Kim: This is CFO of SK Telecom. Thank you for your questions. I would first like to address your question on adding more color on our year-end dividend. As you may already know, SK Telecom announced a new shareholder return policy in April, involving the use of more than 50% of consolidated net income as funds for shareholder return. And for the entire year so far, the DPS per quarter has been determined at 831. As for total shareholder return for the entire year, it very much depends on the closing of the fourth quarter earnings and the BOD deliberation. So I won’t be able to give you a definite answer now, but the company remains committed to the idea that we need to maintain a stable dividend trend as much as possible because dividend is crucial for our corporate valuation.

As you pointed out, in the third quarter, consolidated net income fell about 9% year-over-year due to the recognition of valuation losses from investment assets as nonoperating expenses, but I would like to make it very clear that these valuation losses are one-off nonrecurring items, so they will not be included when calculating funds for dividend payout. Now I’d like to address your question on our company’s value-up disclosure. We reported the corporate value-up program to the Board on October 24 and made the disclosure. This is the first year when companies are encouraged to disclose their value-up programs. In consideration of the purpose of this program, we prepared our value-up plan based on a clear analysis of the current status and presentation of the future vision so that the company value can be properly acknowledged by the investors.

We included ROE, shareholder return ratio and AI Vision 2030 as key indicators. Our ROE is around 9%, already the highest in the industry, but we aim to further improve it through operational improvement and AI transformation. And for shareholder return, we will maintain the current shareholder return policy of using more than 50% of adjusted consolidated net income for dividend payout, which is already the top level in the Korean stock market. And also through AI Vision 2030, we aim to demonstrate our commitment to future growth beyond telco that proposing the total revenue target of KRW 30 trillion and the AI portion of 35% by 2030. By making a report to the BOD, we demonstrated our strong commitment and enhance the plan’s credibility. And going forward, we will develop specific action plans and conduct regular monitoring on implementation, and we will actively communicate with the market.

As you mentioned, no telcos, including SK Telecom, were included in the Korea Value-up Index announced by the Korean Exchange on September 24, but we will strive to enhance corporate value and actively communicate with the market so that we can be included in the index in the future. Thank you.

Operator: [Foreign Language] The following question will be presented by Soojin Kim from Mirae Asset Securities.

Soojin Kim: I would like to ask 2 questions. As you mentioned already, SK Telecom has been pursuing diverse businesses regarding and using AI. And I’d like to understand AI’s contribution to your earnings and also the directions for your AI service development and business models. The second question has to do with the Perplexity Pro promotion. And I understand that this promotion has been receiving positive response in the market. So what kind of benefits did you reap from this promotion, maybe in terms of the increase in the number of subscribers to A. or contributions to the earnings overall?

Yang Seob Kim: Thank you for your questions. Let me first address your question on our monetization plans using AI. For us, under the AI Pyramid Strategy, SK Telecom is developing concrete monetization plans, focusing on AI DC, AI B2B and AI B2C as the 3 main domains to generate profit using AI. First of all, AI data center is a key infrastructure for AI, and it is an area that we can pursue first. Based on the partnership agreement with Lambda that was signed in the second quarter, we began to deploy NVIDIA’s GPUs in Gasan Center. And when Lambda AI data center is opened in December, as scheduled, we will be able to provide the GPU as a Service, which is an AI cloud service. In addition, we will combine energy solutions to GPU as a Service through cooperation with the group to secure and validate solutions optimized for AI data centers and pursue global expansion.

Second, AI B2B market is about selling AI solutions for enterprise customers, which has a very high growth potential. And for this market, in addition to cooperation with big tech companies such as Entropic and OpenAI that the company has been pursuing, A.X that we have developed on our own is also becoming increasingly important in terms of security and cost. So for AI enterprise business, we will pursue mutual cooperation at the group level to satisfy AI needs of group affiliate companies as well as corporate clients and further expand globally. Finally, for the third domain, AI B2C, SK Telecom is indeed the first company to enter the AI B2C market with A.. The app overhaul in the second quarter and the introduction of Perplexity have received positive responses in the market, and we were able to confirm domestic customers’ strong interest in AI.

Starting with the launch of A. phone, we will expand A. agents to more diverse external service domains, and we will be able to offer truly meaningful AI agent experience based on a deeper understanding of customers in diverse context. And we will test the global marketability of global TAA through a closed beta testing within this year. And based on diverse service offerings both at home and abroad, we plan to explore monetization potentials of AI services going forward. Thank you. Now on your second question on the promotion with Perplexity, the Head of AI growth strategy will respond.

A – Unidentified Company Representative: I am [indiscernible] Head of AI Growth Strategy. We carried out the promotion for 2 months from September 4 for the Pro model of Perplexity, which was announced at the time of SK Telecom’s investment of $10 million in Perplexity in the second quarter. In order to provide global level AI services to our customers, we offer 1 year free subscription to Perplexity Pro model, which is worth about $20 per month to SK Telecom customers. Well, in the past, Perplexity carried out similar promotions in other countries such as Japan and Germany. And in Korea, thanks to the marketing capabilities and our channel competitiveness of SK Telecom, this promotion was able to garner the largest number of customers in the shortest amount of time.

Thanks to this unprecedented level of success, SK Telecom has been able to pursue and forge partnership with Perplexity. And currently, our 2 companies are collaborating very closely in terms of technology and marketing. The 2 companies are currently working on developing AI products. And when the promotion is over, we will be able to offer Perplexity Pro model subscription to all customers, not just SK Telecom’s customers, to all Korean people at much more favorable terms and conditions. Now in terms of cooperation on technology, SK Telecom and Perplexity are jointly developing Korean-enhanced AI search model by combining Perplexity’s advanced AI search technology and SK Telecom’s AI assets as well as technology so that this engine can take into consideration Korean-specific cultural context and provide responses that are tailored to the Korean users.

This promotion has demonstrated our capabilities in terms of marketing and SK Telecom’s assets and technology as an AI company. We have both data, multimodal data as well as A.X LLM, AI data centers and GPU farms. So you can see how we have capabilities across the entire value chain. In addition to our cooperation with Perplexity, we will continue to cooperate with various global AI companies to offer AI experience and choices to our customers and expand the market.

Operator: [Foreign Language] The following question will be presented by [Toni Li] from Equity Securities.

Unidentified Analyst: My first question is the impact of the launch of flagship handset series on your 5G subscriber edition as well as marketing costs. And my second question is to ask you for some updates on your cooperation with Penguin Solutions in Lambda for your AI DC business.

Hee Jun Chung: Thank you for your questions. And regarding your question on marketing expenses, the Head of Integrated Marketing Strategy Office will address.

Kim Ji-hyeong: I’m Kim Ji-hyeong, Head of the Integrated Marketing Strategy Office. In the third quarter, on top of Samsung Galaxy Foldables series launched in July, new iPhone series, which typically were released in the fourth quarter, were released early in September. So we had more diverse flagship handset lineups compared to the other quarters. In addition to the effective flagship handsets, 5G price plans that add popular benefits such as Netflix’s subscription discounts are still very popular, which contributed also to sound earnings. Thanks to these factors, 5G subscriber net adds in the third quarter were 350,000, up 20% Q-o-Q, and the 5G penetration rate is nearly 73%. And we believe that these outcomes have resulted from our focused efforts on services and price plans with differentiated benefits that the company has been emphasizing as well as sophisticated marketing activities based on accurate analysis of customers’ needs and preferences.

Going forward, we will continue to implement this type of marketing strategy to generate strong revenue and improve profit in MNO business in the mature 5G market. Thank you.

Hee Jun Chung: Next, I’d like to invite the Head of Enterprise Business Strategy Office to respond to your question on AI data centers.

Unidentified Company Representative: This is [indiscernible], Head of the Enterprise Business Strategy Office. Let me answer your question on our AI data center business. Of the AI Pyramid Strategy in the AI infrastructure domain, with the rise of generative AI, there is a big growth in demand for AI data centers. Against this backdrop, we aim to lead the global AI DC ecosystem through partnerships with Lambda and Penguin Solutions. First of all, in December this year, we will open the first AI data center in the Asia Pacific region, together with Lambda. And we plan to deploy thousands of GPUs in the next 3 years to provide an optimal environment for high-density GPU server operation. And we plan to use the data center to start a subscription-based AI cloud model, which is the GPU as a Service business.

Our core competitiveness lies in our own technological capabilities as well as partnerships with global AI companies. Through cooperation with Penguin Solutions that has expertise in large-scale AI cluster design, implementation and operation, we are strengthening AI DC total solution capabilities, which are going to play a key role in securing leadership in the global AI data center market. Also, in terms of AI revenue generation, among the 3 domains of AI value chain, we expect the most rapid revenue generation in AI data centers and GPU as a Service business will be the first step. In the rapidly changing AI market environment, we also plan to introduce various technology for efficient AI data center operation. In particular, we are considering ways to reduce power consumption and operational costs by utilizing DSIM, an immersion cooling technology.

We’re currently conducting a proof of concept in Bundang. These technologies are expected to further improve cost efficiency.

Operator: [Foreign Language] The following question will be provided by Hoi Jae Kim from Daishin Securities.

Hoi Jae Kim: I’m Kim Hoi Jae from Daishin Securities. My question has a little bit of overlap with the previous question. In your value-up plan, you presented the 2030 targets, so revenue of KRW 30 trillion and the AI portion of 35%. So I’d like to understand the portion of AI in your total revenue as of today. And what is your expectation going forward?

Yang Seob Kim: Thank you for your question. I’d like to comment on our 2030 targets. For the past 3 years, SK Telecom has been actively pursuing the AI Pyramid Strategy. And when it comes to areas for AI monetization, we’ve concluded that AI DC, AI B2B and AI B2C will be the 3 main areas, and we’ve decided to focus on these areas. Now as for AI DC and infrastructure, AI B2B and AI B2C, in all of these markets, there is a rapid growth in demand. And so SK Telecom aims to develop services as quickly as possible so that we can gain leadership in these rapidly growing markets and continue to become an AI company. First of all, in the AI DC market, there is already demand exceeding supply so we believe that we will be able to generate revenue in the short period — in the near future, and there is much more visibility in monetization in the AI DC market.

And moving on to the AI B2B market, with the development of generative AI, we expect very big growth potential of this particular market. And the B2C market has been led by platform companies, but we are seeing a paradigm shift in this B2C market. This will bring new opportunities for SK Telecom. So our 2030 targets of total revenue of KRW 30 trillion and the AI portion of 35% are indeed ambitious targets. But please consider that our value-up plan demonstrates our strong to increasing the portion of AI to the level of the current telecommunications business so that we can transform into a global AI company. And as for the breakdown of revenue between telecommunications and AI, we are working on this, and we will soon be able to communicate with the market when details are ready.

Hee Jun Chung: With this, we would like to conclude the earnings conference call for the third quarter of 2024 for SK Telecom. If you need further information or have additional inquiries, please feel free to contact the IR team. Thank you.