Hello ladies and gentlemen. Thank you for standing by for GDS Holdings Limited’s first quarter 2024 earnings conference call. At this time, all participants are in listen-only mode. After management’s prepared remarks, there will be a question-and-answer session. Today’s conference call is being recorded. I will now turn the call over to your host, Ms. Laura Chen, Head of Investor Relations for the Company. Please go ahead, Laura.

Hello everyone and welcome to the first quarter 2024 earnings conference call of GDS Holdings Limited. The Company's results were issued via newswire services earlier today and are posted online. A summary presentation, which we will refer to during this conference call, can be viewed and downloaded from our IR website at: investors.gds-services.com. Leading today’s call is Mr. William Huang, GDS’s Founder, Chairman and CEO, who will provide an overview of our business strategy and performance. Mr. Dan Newman, GDS’s CFO, will then review the financial and operating results. Ms. Jamie Khoo, CEO of GDS International, is also available to answer questions. Before we continue, please note that today’s discussion will contain forward-looking statements made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995. Forward-looking statements involve inherent risks and uncertainties. As such, the Company’s results may be materially different from the views expressed today. Further information regarding these and other risks and uncertainties is included in the Company’s prospectus as filed with the U.S. Securities and Exchange Commission. The Company does not assume any obligation to update any forward-looking statements except as required under applicable law. Please also note that GDS’s earnings press release and this conference call include discussions of unaudited GAAP financial information as well as unaudited non-GAAP financial measures. GDS’s press release contains a reconciliation of the unaudited non-GAAP measures to the unaudited most directly comparable GAAP measures. I will now turn the call over to GDS’s Founder, Chairman and CEO, William Huang. Please go ahead, William.

Hello, everyone. This is William. Thank you for joining us on today's call. The top priority of GDS senior management is to create value for our shareholders and drive share price recovery. Our business now has two distinct segments, China and International. For China, we believe that the key to creating shareholder value is: First, to get back on to a higher growth track in terms of EBITDA. Second, to generate positive free cash flow before financing and reduce debt. And third, to position strategically for the coming AI wave. For International, the Series A capital raising set a benchmark of nearly $4 per GDS share. We believe that this value will appreciate significantly as we build on our initial success. Now, let’s review our progress towards these goals in more detail, starting with China on Slide 5. The key to restoring higher growth in China is the move-in rate. Over the past couple of years, we focused our sales efforts on opportunities with faster move-in schedules and reasonable pricing. Even though the market as a whole slowed down, we made good progress with winning this kind of business. The results of our efforts are now starting to become visible in our Gross Additional Area Utilized. In 1Q24, the gross move-in for China was 17,000 square meters, all of which was in Tier 1 markets. It’s the highest since 2020. Going forward, based on contractual commitments in the backlog, we expect gross move-in to continue at these higher levels. From the beginning of 1Q24, we stopped recognizing revenue and deducted 12,000 square meters from Area Utilized for three B-O-T data centers which we plan to transfer to the customer on an accelerated basis. During 1Q24, there was around 6,000 square meters of customer churn, most of which we immediately replaced with new customer commitments in our 1Q24 bookings. Over the next couple of quarters, we expect the impact of these one-time factors to diminish. As a result, Net Additional Area Utilized in China will step up in line with the improved gross move-in. Turning to Slide 7. How do we achieve steady EBITDA growth while at the same time generating positive free cash flow before financing? The key is to increase utilization of existing assets and to only incur additional capex when needed to deliver capacity to customers with confirmed move-in schedules. In 1Q24, we brought 14,000 square meters of new capacity into service in China at three data centers in Shanghai, Changshu, and Langfang. The commitment rate for these three data centers is 100%. By the end of the quarter, the utilization rate was already over 40%. This is the pattern we are aiming for. In the past couple of years, we put the brakes on our development program in China, completing around 30,000 square meters of projects per annum in 2022 and 2023. In the current year, we expect a higher level of completions, at around 60,000 square meters, due to the higher level of customer move-in. However, to deliver this capacity, we only need to incur the “cost to complete”, which works out at around Rmb 25 million or less than US$3 million per megawatt. As you can see on Slide 8, we are in a position to grow our area utilized by over 50% while only needing to incur cost to complete of around Rmb 7.4 billion. Turning to our sales on Slide 9. In the current market, we have been selectively targeting opportunities which fit our capacity and have the right commercial terms. In 1Q24, new bookings in China were around 9,000 square meters, most of which relates to inventory at data centers in service. So far, there has not been a lot of AI-driven demand in Tier 1 markets. However, there have been AI deployments in remote locations that are not our focus. These deployments are mainly for AI development rather than AI-enabled applications. Nonetheless, it is an encouraging lead indicator of latency-sensitive AI demand coming to Tier 1 markets in future. As we have seen in our International business, AI requires unprecedented scale and fast delivery. We are very well-placed to satisfy this kind of requirement in China’s Tier 1 markets because of the land and power we have secured at multiple sites. We will use this resource very strategically to capture the AI wave. Turning to International on Slide 12. Our International strategy is based on: anticipating new waves of demand and evolving requirements; moving decisively to secure land and power with short time to market; winning game-changing customer orders; leveraging our competitive strengths to execute faster and more efficiently; and financing the business on a standalone basis. Within a few years of launching our International strategy, we are well on the way to developing a market-leading presence in three of the world’s largest data center hubs, namely, Singapore-Johor-Batam, Hong Kong, and Tokyo. Across these three hubs, we currently have 75 megawatts in service, 196 megawatts under construction, and over 500 megawatts of land and power supply held for future development. Subject to demand, all of this capacity could be constructed and delivered within 3-4 years, which is a critical consideration for customers. We currently have 182 megawatts of commitments, with around 40% from leading global customers. In 1Q24, we won an 18 megawatt order from a local cloud player and, in the current quarter, we won a 43 megawatt order from a global cloud service provider. Both of these orders were for our two Johor campuses. Our pipeline of new business for Johor is exceptionally strong and I am pleased to report that we are in the process of contracting our first business for Batam. Meanwhile, in Hong Kong, we already sold out our first two data centers. In Tokyo, we are partnering for two new data centers, which we believe will be highly marketable. In today’s markets, it is very typical for customers to require a short lead time to delivery of only a few quarters, and they then commit to rapid move-in. Our ability to meet these requirements sets us apart. As a proof point, we have already delivered 70 megawatts in Johor which was 100% revenue-generating by the end of 1Q24. We have another 87 megawatts backlog in Johor, most of which is scheduled for delivery and will become revenue-generating over the next six quarters. Due to accelerated sales pipeline and strong investor demand, we decided to upsize the Series A new issue by US$85 million to US$672 million. This successful new issue demonstrates our ability to access capital for International on a standalone basis. We have now established a channel for future capital raises and further value benchmarks. I will now pass on to Dan for the financial and operating review.

Turning to Slide 15. From 2Q24, we will start to provide segment reporting in our earnings release. As you can see, we have already included most of the segment information in our 1Q24 earnings presentation. We will define two segments: DigitalLand Holdings Limited and its subsidiaries, which comprises all of our business and assets outside of Mainland China, except for some minor third-party data centers in Hong Kong, will be referred to as GDSI or International. GDS Holdings Limited and all of its subsidiaries excluding GDSI, which comprises our ultimate holding company and all of our business and assets in Mainland China, will be referred to as GDSH or China. Turning to Slide 16. In 1Q24, consolidated revenue increased by 9.1% and Adjusted EBITDA increased by 4.7% year on year. Starting with the China segment, in 1Q24 GDSH revenue increased by 1.8% and Adjusted EBITDA decreased by 1.6% year on year. Without the B-O-T transfers, GDSH revenue would have increased by 3.4% and GDSH Adjusted EBITDA would have increased by 1.4% year on year. GDSH revenue growth was mainly driven by an increase in total area utilized of 7.5% year on year, offset by reduction in MSR. GDSH Adjusted EBITDA growth was further impacted by higher power tariffs, which resulted in a decrease in GDSH Adjusted EBITDA margin from 48.6% in 1Q23 to 46.9% in 1Q24. In 1Q24, net additional area utilized for China, before the B-O-T transfers, was 10,858 square meters, which is slightly higher than the average for the prior four quarters. Looking forward, we expect net additional area utilized for China to step up over the next few quarters as a result of higher gross move-in and reduced impact from one-time factors. We also expect the reduction in MSR to slow down and, assuming that power tariffs remain at current levels, we expect GDSH Adjusted EBITDA margin to stabilize, with just the usual seasonal fluctuations. Turning to International, GDSI recorded strong revenue and adjusted EBITDA growth as its first data centers entered service and began to ramp-up with nearly 20,000 square meters of net additional area utilized in a single quarter. Because of the scheduled delivery and move-in commitments, we expect the numbers for GDSI to increase rapidly. Turning to Slide 19. In 1Q24, our China capex totaled Rmb 894 million. Capex during the first quarter is usually elevated as payables are settled on an accelerated basis before Chinese New Year. We expect lower capex per quarter over the rest of the year and still maintain our Rmb 2.5 billion guidance for China capex for the full year. In 1Q24, our International capex was around Rmb 702 million. As William mentioned, we have an 87 megawatt backlog to deliver over the next six quarters. In addition, we plan to purchase additional land in Johor and Singapore and to commence new projects as we win customer commitments. Our capex guidance for International in 2024 is Rmb 4.0 billion. Based on the strong sales pipeline, International capex may accelerate over the next few quarters. Turning to Slide 20. Cash flow before financing for the China segment has fluctuated between positive and negative for the past five quarters. It was negative in 1Q24 due to slower collections and faster payments, which follows the same pattern for the past three years. We still expect to be close to breakeven for the full year. We expect to receive proceeds from the B-O-T transfer in the second or third quarter. In 1Q24, International on a standalone basis had negative cash flow before financing of over Rmb 730 million. With the proceeds of Series A, we have enough capital to complete the current projects. Turning to Slide 22. On 26 March 2024, we announced that we had entered into definitive agreements with certain private equity investors to subscribe for US$587 million of Series A convertible preferred shares newly issued by GDSI. On 13 May 2024, we entered into amendments to the definitive agreements, which included increasing the size of the Series A new issue to US$672 million at the same pre-money equity valuation of US$750 million. We expect the Series A new issue to close on 4 June 2024. Post closing and on an as-converted basis, GDSH will own approximately 52.7% of the equity interest of GDSI in the form of ordinary shares. The remaining 47.3% equity interest will be held in the form of Series A shares by the private equity investors. Turning to Slide 23. The proceeds of the equity capital raised by GDSI are ring-fenced. We therefore believe that it makes more sense to look at our leverage on a segment basis. After closing of Series A, GDSI will repay all shareholder loans and other amounts due to GDSH. At the end of 1Q24, this totaled Rmb 1.7 billion. On a pro forma basis, the cash balance of GDSH will increase to Rmb 9.0 billion, all of which is available to support the China business. The net debt to LQA Adjusted EBITDA multiple for GDSH was 7.7 times. This calculation does not take into account the value of GDSH’s equity interest in GDSI. Turning to Slide 24. During the period from 2Q24 to 4Q24, we have Rmb 1.7 billion of project loan amortization for China. We continue to successfully refinance GDSH onshore project loans, extending maturity and lowering cost. We are also able to draw down on existing project loan facilities to finance a substantial part of GDSH incremental capex. As you can see in the loan maturity schedule, GDSI has obtained 5-year project term loans to finance its developments. Turning to Slide 25. We are not changing our formal guidance for FY24 revenue, Adjusted EBITDA, and capex. We would now like to open the call to questions.

Thank you. We will now go ahead with our first question from Jonathan Atkin from RBC Capital Markets. Please proceed.

Thanks. I got one question around China domestic business and then maybe one international, if I could throw that in. Inside of China, what are you seeing apart from the utilization rate on a square meter basis that you have reported? What are you seeing with respect to power draw and customer behavior around increasing the draw of power that they're contractually able to utilize? Any trend there that might be instructive in terms of increasing demand or maybe follow on demand from customers? And then my question on internationals is that the GDSI, it appears they have a lot of project loans, and I wonder if you could provide a little bit of color on cost of capital and just the sort of counterparties that you have for these loans. Are they domestic, international, etc.? Thank you.

Yes, Jon, it's Dan. I'll address your questions. Regarding the power draw in China, most of our established data centers, which cater to large cloud and internet customers, are operating at their maximum power levels. The total power capacity available in these data centers is fully committed to our customers, who run their businesses with a high level of efficiency. Consequently, we do not have any unutilized or unmonetized power capacity. For new data center developments, we generally build at a higher power density, often exceeding 3 kilowatts per square meter, which may relate to your inquiry. As for project loans in our international business, we are applying a similar strategy as we did with our China business, allocating capital on a project-by-project basis and using local debt to leverage that. The customer contracts, mainly signed in Malaysia, are priced in either US dollars or Malaysian ringgit, so our revenue is in both currencies. Currently, we are borrowing in Malaysian ringgit to minimize our foreign exchange exposure. The loans are with a group of banks that know us well, and we plan to follow a similar process as in China by establishing a structure, building relationships with local banks, and transitioning to having mainly local bank partnerships over time.

Thank you. And then for Japan, the 36 megawatts, can you give us a sense of when you would intend to start construction?

Jonathan, this is Jamie. So on the Japan side, our partner, which is Gaw Capital, will be doing up the construction of the core and shell. So that will be completed and passed on to us by 2025, or early 2026. So then we will start our M&E construction and that will bring us to Q4 2026 for delivery.

Thank you. We will now take our next question. This is from the line of Yang Liu from Morgan Stanley. Please go ahead.

Thanks for the opportunity. I have two questions here. The first question is regarding the asset monetization in China because previously management mentioned about this strategy. What are the opportunities you are seeing in the market and what is the current plan? Or what should we expect on this front? And should we think the transfer of B-O-T is part of that or not? That is the first question.

Yang, let me answer that question, then you ask your next question. First of all, on the B-O-T transfer, this is very specific. We have 15 B-O-T data centers. This transfer involves three of them. One of the three is at a campus where it is the only data center that we have invested in and operate. So that is being transferred for the sake of operational efficiency. The other two data centers, the customer has a change in plan in terms of how they wish to utilize those data centers, which includes the way in which they are fitted out. But this is not a change of strategy. It's not something that we expect to happen in the future. We came to a mutual agreement. We will recover our investment plus a reasonable return over the period of time in which our capital has been invested. And it will make a small positive contribution to our cash flow before financing when the proceeds are received, either in this quarter or next quarter. They talked more generally about asset monetization. Yes, I appreciate we have talked about that for some time, and it is most definitely a strategic objective of ours. And I think that we are moving in the right direction. We have not ceased to make efforts, and currently, we have a number of projects ongoing, including one end of the spectrum, C REIT or China REIT. One degree over from that is what's referred to in China as a private REIT, which involves exactly the same structure as a public REIT, but doesn't have the public REIT at the top of it, but is a stepping stone in terms of monetizing an asset which can then subsequently be injected into a public REIT. We also have other structures that are more like financing, and we're dealing with China's leading insurance companies, leading RMB and private equity funds, and even some US funds who are looking at assets in China. I think we're very determined about this, and I think there's a chance we get something done before the end of this year because it's not in any of the numbers or guidance that we've provided, but clearly, it would make a contribution to our free cash flow before financing, and I'm quite sure it will be accretive.

Thank you. Yes, I have another question. In terms of the CapEx outlook beyond 2024, do you think the China part of CapEx can further come down next year if the demand stays at current level or if the gross move-in stays at current pretty good run rate? Thank you.

Our business plan anticipates an increase in our growth rate primarily due to the contracts within our backlog. We are not making assumptions about broader market trends; instead, we are focusing on what we have already secured and are working to deliver. We expect the move-in to reach a high level and maintain that level for the foreseeable future. The capital expenditure guidance for this year is RMB2.5 billion. While it's early to provide future guidance, our business plan suggests that capital expenditures in the next one or two years will be around that amount or lower.

Thank you. We will now take our next question. This is from Frank Louthan from Raymond James. Please go ahead.

Hey, this is Rob stepping in for Frank. You may have mentioned this earlier, but what is the effect of rising interest rates on your customers' businesses? How should we view that impact moving forward?

In China, the trend for interest rates has been contrary to what we observe in the US and many developed markets. The reference interest rate, known as the over five-year loan prime rate, is at its lowest since we began operations. Additionally, the margin that banks apply to our project financing has decreased, now averaging just a few basis points above or sometimes several tens of basis points below the loan prime rate. Consequently, debt financing costs in China are the lowest they have ever been. Our primary customers are large cloud and Internet companies, most of which have no debt, so I don't believe this situation significantly impacts their business either way.

Thank you. We will now take the next question. This is from the line of Cooper Elias from TD Cowen. Please go ahead.

Hi, everyone. You have Cooper Belanger on here for Michael Elias. I wanted to ask a quick question regarding the segmentation. Obviously, you provide guidance for GDSI and GDSH CapEx separately. Should we expect the same thing going forward in terms of revenue, adjusted EBITDA, etc.?

Thank you, Cooper. For the time being, the answer is not in a formal sense, but during the prepared remarks, we will continuously update and give some direction on the key performance indicators, both operating and financial KPIs. Maybe after a few quarters, we might revisit that. But for now, I think we split out on a historical basis all the numbers that really matter. I think the only one which we have not split is MSR because for now, it's not material to look at China and international MSR because there's not enough difference. But when there is, we will split that out, and then we will provide commentary on each of these metrics on a China and international basis. So I think that will probably get you a long way until we provide formal guidance for GDSH and GDSI separately.

Thank you. And we now have a follow-up question. This is from the line of Yang Liu from Morgan Stanley. Please go ahead.

Thank you for the chance to ask a question again. I would like to start with the demand side. Do you notice that the demand from Chinese customers is improving compared to three or six months ago? We observed that major Internet companies increased their capital expenditures significantly in the recent quarter. Is this reflected in the demand for GDS? Additionally, we previously saw strong overseas demand, and I'm looking for an update on that in comparison to three months ago. Is it showing improvement or has it moderated a bit? I also want to ask about the recent increase in private equity financing, which means that GDS Holdings' stake in GDSI will decrease to around 52%. Does GDS have a strategy to consolidate GDSI in the long run? Would you be comfortable if future financing rounds result in your stake falling below 50%? Lastly, do you have concrete plans to spin off GDSI? Thank you.

I believe I've addressed the demand question. In China, we are beginning to see demand recover. This is evident from our first-quarter move-in rate, which is the highest it has been since 2020, making this the strongest quarter in the last four years. This aligns with our expectations, and we anticipate benefiting from it. In terms of new incremental demand, we've observed some growth, primarily driven by AI, possibly for training purposes. Some deals are directed towards remote areas, which are not our primary focus. However, we expect demand to increase for inference or AI-enabled applications, which will positively impact us in the near future. Overall, demand is starting to recover, and our goal is to accelerate customer move-ins, which we expect will continue to occur in China. On the international front, demand is noticeably stronger than it was three months ago. We are seeing a diverse range of multinational customers reaching out to us for resources in Johor and Batam, and the size of deals is also increasing. We are very confident in securing more deals in the next 12 months. In Japan, following our recent announcement, we have seen a surge in sales leads. If we can deliver by the end of 2026, we expect to have some pre-sales in Japan.

From GDS Holdings' perspective, our primary goal is to ensure GDS International achieves maximum success and that the value of our investment in GDS International increases significantly. To optimize GDS International's success, it is very likely that they will seek additional capital funding. We have already noted how such actions can impact our ownership stake and our capacity to consolidate. This was somewhat anticipated in the design of the Series A new issue, where we included specific rights to safeguard GDS Holdings' interests while also allowing us the option to initiate an IPO and distribute shares to our shareholders in the future. Our focus is on maximizing the value for our shareholders, ensuring that any increase in the value of our international investment translates into benefits for them. If this value isn't reflected in our share price, we will need to explore alternative methods to ensure it benefits our shareholders. This is why we possess the unique right to make future decisions about pursuing that strategy.

Thanks. Quite encouraging to hear the plan. Thank you.

Thank you. And we will now take our next question. This is from Gokul Hariharan from JPMC. Please go ahead.

Hi. Thanks for taking my question. William, you did talk about some of the initial AI demand that is starting to show up in China, especially for training. Could you talk a little bit about what kind of data center capacity or power profile that you need to prepare? Are there distinct differences in terms of the kind of data centers required for AI workloads that you are hearing from your customers compared to the regular cloud data centers that you always had? And do you feel that you will have to start increasing to build some of these data centers eventually, or you don't think that you can accommodate them in the existing data centers themselves?

I believe the AIDC depends on how you define it. If you consider the data center hosting the GPU as AIDC, we are already there. Typically, our existing data centers in Tier 1 markets or in the edge towns of those markets have sufficient power capacity to support high-density servers, a configuration we established a few years back. Over the last five or six years, our new data centers have already been constructed with high power density in mind. Regarding the differences, while it may not be a trend right now, AI data centers in China may eventually need more cooling systems, possibly air cooling, to manage their cooling requirements. However, we have experience with this since we've built liquid cooling type data centers for two of our largest customers four years ago. AI data centers are nothing new for us, and we have established several similar liquid cooling data centers in Johor since last year.

Hello?

Yes, we're still here.

I apologize for missing the end of William's answer. Moving on, my next question is about the international business. Please continue, William.

Yes, Gokul, I think my conclusion is that we started to implement it a couple of years ago for the AI data center. This is meaning now everybody calls it an AI data center, right? So, we already have most of our Tier 1 market data center suite for this kind of requirement.

Got it. My second question is on international, maybe to William and Jamie. What are you seeing in terms of the local competitors, especially in Malaysia and to some extent, we are also seeing in Indonesia, there's a lot of announcements coming through from local competitors, coming from the data center industry, coming from other utility industries as well. Are you starting to see them in some of the bids that you're participating in, or is this still a separate market for you compared to these local players who are announcing big data center deals?

So I think in Malaysia typically, let's say Malaysia or Indonesia, Batam, right, I think we definitely have the first movement advantage, number one. I think because we are the pioneer to step in this market and because we already know, we know better than anyone else in this region about the technology trend, and we know we are much better understanding our customer needs, so we know where they will go and when they will go and how we will do. So, I think this is a different advantage we already have than anyone else in this market. So, frankly speaking, before 2028, I think most of the power in this region, we already secured most of the power. So I think even a lot of new players jump into this market, I think the time to market will be way behind us.

Got it. Yes. Thank you very much. Thanks. Thank you. We will now take our next question. Please stand by. Next question is from the line of Sara Wang from UBS. Please go ahead.

Thank you. Just one quick question. So, for the China business, given the backlog might be signed a couple of years ago. So I'm just wondering if the AI-driven demand will simply drive acceleration of execution of the previous backlog, or will there be any changes to the contract terms signed maybe previously. Thank you.

Thank you. Your line is open.

Yes, I believe the main factor in our move-in growth in China is the influence of the Internet and AI, despite traditional growth being quite slow. Currently, the majority of demand stems from AI and Internet companies. As I mentioned, we are already starting to see the benefits of this trend. Our first-quarter move-ins are primarily driven by these AI and Internet firms, and we anticipate that this will continue.

Yes, let me have a go. Obviously, our backlog is mainly cloud service providers. Yes. So, I listen to China's largest cloud service providers' earnings call, and they talk about how they're integrating AI into their cloud business and how AI development is driving demand for their cloud business. So I think if you have cloud service providers in your backlog, then yes, their successful development and bringing to market AI-enabled products and services will contribute to, you could call it AI growth, but also to demand for cloud services.

Thank you. As there are no further questions, I'd like to now turn the call back over to the company for closing remarks.

Thank you all once again for joining us today. If you have further questions, please feel free to contact GDS Investor Relations through the contact information on our website, or the Piacente Financial Communications. See you next time.

This concludes this conference call. You may now disconnect your line. Thank you.