© 2024 Alpha Spread Limited. All Rights Reserved.
www.alphaspread.com is not operated by a broker, a dealer, or a registered investment adviser. Under no circumstances does any information posted on www.alphaspread.com represent a recommendation to buy or sell a security. The information on this site, and in its related newsletters, is not intended to be, nor does it constitute, investment advice or recommendations. The information on this site is in no way guaranteed for completeness, accuracy or in any other way. In no event shall Alpha Spread Limited be liable to any member, guest or third party for any damages of any kind arising out of the use of any content or other material published or available on www.alphaspread.com, or relating to the use of, or inability to use, www.alphaspread.com or any content, including, without limitation, any investment losses, lost profits, lost opportunity, special, incidental, indirect, consequential or punitive damages. Past performance is a poor indicator of future performance.
You don't have any saved screeners yet
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Hello, ladies and gentlemen, thank you for standing by for GDS Holdings Limited's First Quarter 2022 Earnings Conference Call. [Operator Instructions] Today's conference call is being recorded. I'd now like to turn the call over to your host, Ms. Laura Chen, Head of Investor Relations for the company. Please go ahead, Laura.
Thank you. Hello, everyone. Welcome to the first quarter 2022 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 investorsgdsservices.com.
Leading today's call is Mr. William Huang, GDS Founder, Chairman and CEO, who will provide an overview of our business strategy and performance; Mr. Dan Newman, GDS CFO, will then review the financial and operating results; Ms. Jamie Khoo, our COO, 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 perspective as filed with U.S. SEC. The company does not assume any obligation to update any forward-looking statements, except as required under applicable law.
Please note that GDS earnings press release and this conference call includes discussions of unaudited GAAP financial information as well as unaudited non-GAAP financial measures. GDS 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 Founder and CEO, William. Please go ahead, William.
Thank you. Hello, everyone. This is William. Thank you for joining us on today's call. We are operating in a difficult environment. Everyone is having to deal with unprecedented challenges, including the recent COVID lockdown in China. GDS business is resilient and defensive. We generate recurring revenues, underpinned by long-term contracts with high-quality customers.
Despite the challenges, I'm pleased to report a solid set of results for the first quarter. We grew revenue by 32% and adjusted EBITDA by 29% year-on-year. We continued to win new business with 18,000 square meters of net additional customer commitments. We strengthened our funding position by raising USD 620 million from a convertible bond issued to strategic investors and completed a further USD 530 million of project financing.
In the near term, we must deal with the challenges to the best of our ability. However, at the same time, we'll remain focused on strengthening our strategic position for the medium and longer term through further developing our customer franchise, adding to our resource pipeline, consolidating the market as opportunities arise and accelerating our regional expansion.
With scarce resource already secured, great customer relationships and a proven track record, we have locked in enormous growth potential, which will materialize in the future. It is just a matter of time.
So turning to Slide 4. Starting from the late February, massive lockdowns happened in many areas of China because of the COVID outbreak. We have a total of 82 data centers in service, out of which 58 have gone through lockdown situations, and over 30 are still locked down today. During these lockdowns, around 880 people were isolated inside our data centers, including 660 GDS employees and 220 from our customers.
We needed to keep our data centers in continuous operation, while keeping the people inside safe and healthy. We never failed to deliver. All of our data centers are running as usual without interruption, and all the people inside are taken care of. We really appreciate every effort from our data center employees. It is also highly appreciated and recognized by our customers.
Turning to Slide 5. The first quarter of the year is always a slower season because of the Chinese New Year. On top of this, lockdowns also impacted the move-in rate. Nonetheless, we still achieved over 112,000 square meters of net additional area utilized for the quarter, and our utilization rates increased to 67%.
As shown on Slides 6 and 7, we have always maintained a high commitment rate for our area in service and area under construction. We have a very large backlog totaling 243,000 square meters, which is equivalent 73% of our area utilized. It provided us with high visibility to future growth.
Our backlog is solid. Our data centers are concentrated in Tier 1 market where supply is increasingly scarce. Our customers need this resource. We will continue to deliver the backlog. As I said, it is just a matter of time.
Turning to Slide 8. We have scaled down our capacity delivery this year to align with the current slower environment. In the first quarter of 2022, we brought 4,500 square meters of capacity into service and initiated 1 new data center under construction. SH18 Phase 1, which is 68% backed by an anchor customer commitment.
Turning to the Slide 9 and 10. While delivery is slower for some customers, there are still other customers out there with substantial new requirements. In 1Q '22, we booked 18,000 square meters of new commitments, including 3 hyperscale orders, 2 came from the existing cloud and the large Internet customers, and the remaining one came from a new financial institution customer.
Turning to Slide 11. Continuing the trend which we highlighted last quarter, financial institution and the large enterprises accounted for around 45% of new bookings in 1Q. We are still confident of achieving our full year sales target of around 90,000 square meters net add. From what we see in the pipeline, there could be large contributions from Hong Kong and Southeast Asia than we thought before.
Turning to Slide 12. I have been in Singapore for the past few months, along with our COO, Jamie. We have made substantial progress in our regionalization plan. According to the Cushman & Wakefield, Singapore is a top 5 data center market globally, and it was one of the fastest growing. We know from our whole market customers how much latent demand there is for Singapore. Currently, Singapore government has elected to pursue a moratorium on data center construction. While it may soon allow some new development, the numbers clearly indicate that excess demand will have to go elsewhere.
We're one of the first movers to establish hyperscale Green Data Center projects in close proximity to Singapore. To our understanding, there are no significant legal constraints on the flow of data cross-border between Singapore, Malaysia and Indonesia. Our sites are therefore well-placed to serve both as regional hub in the domestic market. Furthermore, we are the only player to have established projects both in Johor, Malaysia to the north of Singapore, and Batam, Indonesia to the south of Singapore.
The first phase of our 54 MW projects at Noseia Tech Park, Johor is now under construction. To complement this site, we recently signed a partnership with YTL Power to codevelop 168 MW of capacity across an individual design of a facility at a visionary YTO Green Data Center Park, Johor, approximately 30 km from Singapore, this data center will be powered by the on-site solar generation.
We have completed the land purchase in Nusajay -- in Nongsa Digital Park, Batam and the construction of the first phase of our 28-MW project will start in the next couple of months. According to the Cushman & Wakefield, Hong Kong is top 10 data center market globally which offers excellent network connectivity and availability of all major cloud services.
In the first quarter, we signed a built-to-suit lease for HK3, together with our existing projects with HK1, HK2 and HK4. This will give us a continuous supply of high-quality data center capacity over the next 5 years. All concentrated in the favorite West Kowloon area. This is an extraordinary achievement considering how difficult it is to solve for real estate in Hong Kong.
In Macau, we are launching the first ever [ carry multidata ] center project to meet new Internet and digitalization requirements. Across Southeast Asia, Hong Kong and Macau, we now have visibility for over 300 MW of capacity. Our customers are very excited about this unique strategic presence.
We expect to announce several anchor order over the course of this year. Leveraging the strength of our franchise in Mainland China, we believe that within a short period of time, we will create significant additional value for our shareholders through regional expansion.
Now I will hand over to Dan for the financial and operating review. Thank you.
Thank you, William. Starting on Slide 15, where we strip out the contribution from equipment sales and the effect of FX changes.
In 1Q '22, our service revenue grew by 2.6%, underlying adjusted gross profit grew by 2.3% and underlying adjusted EBITDA grew by 2.4% quarter-on-quarter. Our underlying adjusted EBITDA margin was 47.1% compared to 47.2% in the previous quarter.
Turning to Slide 16. Service revenue growth is driven mainly by delivery of the committed backlog and closing of acquisitions. Net additional area utilized during 1Q '22 was 12,545 square meters. Around 7,500 square meters was in Tier 1 markets affected by lockdowns, and 5,000 square meters was in B-O-T projects in unaffected remote areas.
During the second quarter, we expect a similar level of net additional area utilized. Monthly service revenue per square meter was RMB 2,296 per square meter per month, down by 2.4% compared to the previous quarter. It is mainly due to dilution from move-in at B-O-T and edge-of-town projects. For FY '22, we still expect MSR to decline by mid-single digits in percentage terms year-on-year.
Turning to Slide 17. Our underlying adjusted gross profit margin was 52.4% for 1Q '22 compared to 52.5% in the previous quarter. As a result of higher coal prices and a raising of the ceiling on thermal power tariffs in China, we experienced around 10% increase in unit power costs in 1Q '22 as compared to the prior quarter. On a per kilowatt hour basis, we are now paying around 15% more than we were a couple of quarters ago.
Thermal power tariffs appear to have stabilized at the current level. As we indicated on previous calls, we estimate that elevated power tariffs will have around a 1 to 1.5 percentage point impact on our profit margin for this year. However, over time, we expect thermal power tariffs to normalize.
Turning to Slide 18. Our CapEx for 1Q '22 was RMB 4.9 billion, consisting of RMB 2.2 billion for organic CapEx and RMB 2.7 billion for acquisition consideration. As at the end of 1Q '22, we had a liability of around RMB 1.6 billion on our balance sheet in respect of deferred and contingent consideration payable for acquisitions, which had closed by the end of the first quarter. We have a further RMB 463 million for the acquisition of Shenzhen 11, which we announced during 1Q '22 and closed in April. These amounts do not include the cost of buying our partner interest in a few of our projects.
Looking at our financing position on Slide 19. At the end of 1Q '22, we had RMB 11.3 billion or USD 1.8 billion of cash in our balance sheet, and our net debt to last quarter annualized adjusted EBITDA ratio was 7.0x. Our effective interest rate for 1Q '22 was 4.7% compared with 5.4% in the previous quarter or 5.5% for the full year of 2021.
During 1Q '22, we successfully raised USD 620 million through the issue of convertible senior notes with a 0.25% coupon and a 7-year tenor. During 2Q '22, we will reduce working capital loans, which is shown here as short-term debt by around RMB 2.3 billion or USD 350 million.
Turning to Slide 20. As of the end of 1Q, we had total capacity in service and under construction of 660,000 square meters. Against this, we had total area committed by customers of 575,000 square meters. Assuming that we deliver all the backlog and sell out the small amount of remaining inventory, our area utilized or revenue-generating capacity would increase by around 90%.
The total cost to complete all existing projects is around RMB 11.2 billion or USD 1.7 billion, which we can finance with existing resources. It is a relatively small amount of CapEx to generate a large amount of growth because we have already made most of the investment.
On top of our existing projects, we have secured another 460,000 square meters of pipeline held for future development. It's land and buildings with project approvals and energy quota, predominantly located in Tier 1 markets, which we believe is a very valuable asset.
Turning to Slide 21. Based on our performance in 1Q '22 and what we know so far at the current quarter, we are still on track to deliver results within our original guidance range. for revenue and adjusted EBITDA. This assumes progressive relaxation of lockdowns over the next 1 to 2 months, and a moderate pickup in the move-in rate in the second half of the year. We think that this is a reasonable assumption to make at this point in time. Accordingly, we are leaving our original guidance unchanged.
We would now like to open the call to questions. So please, operator?
[Operator Instructions] Our first question comes from Tina Hou at Goldman Sachs.
One question from me is that in terms of the overseas market, especially in Southeast Asia, in the longer term, say, 5 to 10 years, how big of a potential market do you think it could be versus the domestic China market? And what kind of growth rate can we expect from our projects and presence there?
Yes. I'll answer first. I think the Southeast Asian market, mainly driven by the Chinese customer and the U.S. player. And what we see is the huge potential for the local domestic market, because a lot of the unit -- the number of the unit increased very significantly in this region. That will -- that's indicated the future market has a huge potential.
In terms of the population in this area, it's around 600 million population in this region. I think the -- over time, I think the market will grow to a similar level of the Chinese market because, let's say, at least 50%, right, around 50% of the Chinese market. That's my view. But the market is still in the early stage. It's more like 8 years ago in China and -- but the potential is high.
Yes. So in terms of the growth rate, I think the Southeast Asia market is the most fastest growing market in the world based on some analyst report, right? So now the total amount -- demand in current stage compared with China is still small. But we should be well positioned in this region. That's our thinking, yes.
William, can I just have a quick follow-up? So I know our...
By the way, based on my understanding, the current total market in Southeast Asia is 200 MW totally. Existing market, I mean.
Got it. Yes. My follow-up would be is our target customer also including the local and international companies in addition to the domestic Chinese customers? And also, what kind of IRR -- project IRR should we expect for the Southeast Asia project?
In terms of the customer, I think very clear, GDS, our methodology is to serve for all kind of customers in this region. But of course, the first priority is serving our installed base customer, which they have heavy implementation -- the business implementation in this region already. So I think this is our first target. As you said, we are not just serve Chinese customer. We potentially will also talk to all the international player and also local Internet and car player as well.
In terms of the return, Dan, maybe you can add on some color?
Yes. I think the returns will be well up to the levels in China, probably towards the top end of the range of what we typically achieve in China right now. We have a favorable dynamic where the Singapore market is very constrained and there's a very substantial amount of excess demand in the adjacent markets in Malaysia and Indonesia. Development cost, the operating cost, including power tariffs are substantially lower than Singapore.
We already have a very good handle of what the development cost is going to be for us. We also have a pretty good idea of what the selling price is going to be for the initial orders because, of course, we've been in dialogue with potential customers for quite some time. So based on those inputs, we believe that we can offer a very competitive price for this nearshore capacity and still generate returns, as I indicated, which are compared at least on par, if not towards the top end of the range of what we can achieve in China these days.
Our next question comes from John Atkin at RBC Capital Markets.
Two questions. One is, if the CapEx guidance is unchanged, but you talked about a slowing schedule. So what is the kind of the offset there?
And then on M&A, just kind of your view on the M&A environment and how you view financing options going forward in case you needed to write a long-term. [Audio Gap]
Yes. So I'll answer the questions, William. So firstly, just to remind, we gave CapEx guidance a couple of months ago on previous quarterly earnings of RMB 12 billion for this year. And during -- in the script, I gave a breakdown saying that it would be RMB 8 billion organic, of which RMB 6 billion was in China and RMB 2 billion is in Hong Kong and Southeast Asia. RMB 6 billion in Mainland China and RMB 2 billion in Hong Kong and Southeast Asia. And then the balance of RMB 4 billion would be acquisitions and land bank. So the acquisitions and land banking is a bottom-up estimate. So that's deals which have already been done. So that number, there's no change. And out of that RMB 4 billion, we already incurred around RMB 2.7 billion in the first quarter, and we had the balance of payments, acquisition consideration and land purchase payments, which will mainly be made in the second quarter.
For the remaining part, the RMB 8 billion, which is organic in Mainland China and regionally, we only incurred just over RMB 2 billion out of the RMB 8 billion in the first quarter. So that part, you can say, is in line with the kind of quarterly run rate. I mean, 4 quarters at that run rate would add up to RMB 8 billion.
In actual fact, because we've made adjustments to the delivery schedule, which, of course, involves a lot of coordination with customers as well, it takes some time for that to work its way through to CapEx paid. First of all, we have to slow down the incurrence of CapEx, and then it gets reflected in the timing and the amount of CapEx payments. I think that the CapEx in the second half of the year could be somewhat less than it will be in the first half of the year.
On M&A, I think your question, Jon, you're breaking up, but you were saying how would we finance M&A?
Yes. You talked about the RMB 1.7 billion that you need to meet your current development pipeline, but if you wanted to get more aggressive on expansion or through organic growth, how do you view the financing environment?
Yes. Well, if I may, I'll start off by answering that we are fully financed for what we've committed to already. I mean, that's always been our approach. We set aside cash to capitalize the projects with equity, and we secure project debt as early as we can in the life cycle of the -- of each individual projects. And then we have some surplus resources over and above that. If there are opportunities that we wish to pursue, which could be acquisitions or it could be other new business opportunities, then we need to obtain the capital to be able to pursue those.
I think that even in the current environment, we have a very good range of options. Firstly, we have to make a decision of whether to do something through public capital raising or through private capital raising. And then within public and private, there's a range of options. What I would highlight is that since 2016, when we IPO-ed in the U.S., we have done 4 private placements. If you recall, with CyrusOne with Ping An, with Hillhouse and then recently with Sequoia and Sovereign Wealth Fund. That was equity convertible preferred and convertible bonds. We've also done 2 joint ventures, 1 with Sovereign Wealth Fund, 1 with CITIC private equity, the largest private equity firm in China. And just recently, we announced a co-development with YTL. These are all in the category of private capital raisings or accessing capital privately.
So we did this during the 6 years that we've been a public listed company in the U.S. The business fundamentals, we believe, are still very, very solid as we keep emphasizing we have very high visibility for future growth from our asset base and from our backlog. There is very strong interest from private capital providers in investing in this sector, particularly with market leaders. So bottom line of this is that if there are opportunities that we wish to pursue, we will undoubtedly be able to access capital on usable terms to be able to do what we wish to do.
If I could ask one last question then. So the mix of Internet versus enterprise versus cloud inside of China, you talked a little bit about FSI demand and so forth. But maybe just a little bit of an update on what you talked about in the script as well as last quarter.
I don't know if we did want to add, but we talked last time when we talked actually for quite a few quarters about being positioned for where demand comes from. And the demand profile changes over time. So during last year, we saw significantly increased new business from financial institutions, both domestic, Chinese and foreign. We had some tremendous wins, including most of the top 10 global banks and best banks on Wall Street.
In the first quarter, FSI and enterprise business was 45% of new bookings. And in the second quarter, it will probably be a similar level, maybe 40%. So that demonstrates that we positioned our sales effort and our asset base correctly. William, would you like to add something on what we're doing on the FSI enterprise side?
Yes. I think diversify our customer base, always our goal, right, from day one. So that's why let us catch up any kind of growth from the different industry. If we look back last 10 years, we start from the enterprise customer, and then we catch up the Internet booming. Then we go to the cloud take off. And now the demand also more balanced from the cloud, Internet and also the enterprise as well.
So I think we are very, very -- we have a very solid customer base. That's allowed us to capture any kind of the demand from the -- because the market always change. So the only thing -- the same -- the good -- the right thing to do is build -- always build your solid customer base, which we have. So now since last quarter and this quarter, maybe last quarter and last few quarters, I think the growth -- we saw the growth from the enterprise and financial institution. Obviously, the cloud a little bit of slow down, right? But we still can meet our target.
So since last year, we already saw that chance. So last year, we start to hire more -- increase our salespeople to penetrate to the enterprise customer and the financial too. Now we saw the effort is coming, and we did the right thing last year.
[Operator Instructions] Our next question comes from Michael Elias at Cowen.
This is [ Jesse ] on for Michael. I just wanted to touch on the organic leasing target. And you noted in your prepared remarks that you could see a little bit more coming from Southeast Asia in 2022. Wondering if you could provide a little bit of color around the exact split between Southeast Asia from along with Mainland China?
And secondarily, your guidance does imply a step-up in leasing to reach that 90,000 square meter target. What are you hearing from your customers in any case an acceleration is coming on that as part of that, what level of visibility do you have in your existing pipeline for that leasing to come on?
Dan, go ahead.
Yes. I'm sorry, on the first part of the question, we deliberately didn't provide any quantification of a split. We can make our targets in a number of different ways, right? So it's not really appropriate for us to put some kind of quota on orders from a particular type of customer or a particular region. We have a couple of projects in Hong Kong where we haven't yet disclosed anchor customer orders, and we have 3 major projects around Singapore, 3 campuses, 2 in Johor and 1 in Batam. So that's 5 projects in total. Of course, we've made these investments in close consultation with our leading customers. And there has been a sales dialogue going on for a long period of time.
But when we actually disclose the sales order, it depends on there being a firm commitment. It's too hard to say right now how many of these projects will have firm commitments within this financial year. It could be a combination. So that's why it wouldn't really be meaningful or appropriate to put a number on it. I think we will have enough anchor orders in Southeast Asia to give investors a very strong feeling about, call it, proof of concept as to what we're doing. I think that's probably the most important thing.
So your second question was about visibility on the 90,000. William, would you like to add something?
Yes, I think the -- #1, 90,000 square meter, we confirmed we still can achieve, right? We are confident on that. But if you think about -- it's early to talk about -- a little bit early to talk about the accelerate. But in general, I think our customer is still prepared for the future, all our customers preparing for the future growth. So we are monitoring our customer business plan very closely. So in order to well position to catch up whatever it accelerate or still maintain the current demand profile. So I think our goal is long -- mid -- short term, we cannot say what happened, accelerate it -- will it accelerate or not. But still, we are confident for the midterm, long term, right? The demand still will not change.
Our next question comes from Frank Louthan at Raymond James.
I just want to go back to the guidance. So when you came in the end of the year, the guidance was a little bit lower because of supply chain issues. Your customers are having supply chain issues and getting installations. Now obviously, they've been precluded from installing because of the lockdowns. What's the bigger factor now in your mind that if there's something that would keep you from hitting the guidance or have to have you lower it later this year? Are you confident that they do have the equipment and they've gotten through the supply chain issues to be able to do the installations and now is it just the assumption that the lockdowns will pull through? Or how should we think about those 2 factors?
Yes. The main difference is the lockdowns. I think when we gave guidance in mid-March, it was still relatively early. I don't think anyone had anticipated then that, say, Shanghai would be in lockdown for 2 months. And I don't know how much a specific detail investors have about the extent of lockdowns. But what we saw was the lockdown in Shanghai, which is quite well known, but also the surrounding areas where the government, I think, as a precaution went into lockdown at Jiangsu province.
And then there's quite a bit of news about some restrictions in Beijing. But what I don't think is necessarily so well known is that Langfang, which is in Hebei province outside Beijing, that's been in lockdown for extended period of time as well. So if you look at the page in our earnings presentation and it's in the appendix where we show what we call ramping up data centers. Ramping up data centers is where most of the move-in is taking place. So you just scan down that table and you just look at the number of data centers, which start with SH or CS, which is Changshu, or NT, which is Nantong. That's all in the Shanghai area and surrounding the area affected by lockdown.
And then you look at Beijing and Langfang LF, you'll see a very substantial part of our ramping up data centers are in the areas which have been amongst the worst affected by lockdowns. We may be coming to the end of that, right, from -- certainly from government comments. And when we looked at our guidance, we assumed, as I said during the prepared remarks, that lockdowns come to an end progressively in the next 1 to 2 months. And thereafter, we assume that move-in would recover to the level that it was at last year. The last year was not affected by lockdowns, but it was affected by quite a number of other factors. So I think if you look at the quarterly move in last year, excluding acquisitions, it was in the high teens in terms of thousands of square meters per quarter.
So we've assumed that that's what will happen in the second half of the year. And then if that happens, we will come out with revenue [ adjusted ] within our guidance. I think that everyone can probably do their own cross-check mathematically because we've given direction saying that our MSR will decline by mid-single digits, and we have revenue guidance. So if you divide one by the other, you can calculate the average area utilized and then construction kind of quarterly progression for that metric.
Our next question comes from Gokul Hariharan at JPMorgan.
So first of all, on cloud, one of your bigger cloud customers yesterday mentioned that they are becoming a lot more selective in terms of their cloud investments and looking for quality growth rather than growth in any cost. Could you talk a little bit about the discussions you're having with customers, cloud customers, in terms of how they think about growth going forward? Is there a meaningful step down? And what kind of expectations do you have, let's say, in the next 3, 4 years in terms of cloud growth?
And secondly, on the FSI, financial services customers, is there any different profile in terms of financial service customers in terms of the kind of demand that they have on the engagement models that you have? If that becomes a significant part, much bigger part of your revenue, does that mean anything for IRR or MSR?
Yes. I think the cloud service provider still is a major driver -- main driver to drive the data center demand and the growth. I think the -- even they talk about being selective, they still pursue the growth -- the high growth rate. So I think this does not mean they will slowdown, let's say, dramatically. So I think still, they maintain the -- I think the 20%, 30% growth still, I think, is the number they pursue at least. So this is still a big number, right, in my view.
But on the other hand, since we talked about the demand profile a little bit changed because a lot of the enterprise or Internet giants, they start to do their hybrid cloud model. So the demand from the cloud a little bit shifted to the -- our end users to directly purchase data center. So I think the -- in total, from our perspective, the market total demand still maintained a similar size compared with last year. But the demand shifted from one industry to another industry. That's the current trend that we saw.
In terms of the enterprise customer and the financial institution, I think they are not traditional colo demand, a small size. They also represent very big demand, let's say, 1,000 reps, 2,000 reps, a lot of because they changed their IT structure from the -- typical from the mainframe to the server base. So the demand profile for each order still very, very big. You can call, it's like a semi-hyperscale data center -- - data demand, right? So in terms of price Dan, you want to add some color?
Semi-hyperscale more or else answers the question about price as well, actually. It's almost entirely. I'm trying to think of it, it is entirely actually what we would call downtown data centers. So unless we had continued to pursue the strategy of sourcing data centers within Beijing, within Shanghai, which we have done -- I'm sorry, within Shenzhen, which is very time consuming and challenging. Unless we've done that, we would not have been able to get this business. So it's mostly downtown.
There is still a significant differential between the pricing for downtown and edge of town, which I think all customers recognize and accept. I think the IRRs are probably quite similar now between downtown and edge of town. So I don't think this FSI business makes much difference to our average IRRs or to our operating cost structure and so on.
Our next question comes from Joel Ying at Nomura.
Actually, I actually saw a tightening policy trend for the PUE monitoring and cover emission quota management in Beijing area. So should we expect any impact to the company or industry into mid-, long-term? Maybe some old data center, they need renovation. So we might need some time or might be extra capital to solve such issues maybe in that one. Can management talk about that?
So pardon, what's your question?
Yes. Yes. William, I think it was yesterday, it was at the municipal government NDRC announced implementation of this year's monitoring of PUE levels and list out like 8 or 10 actions that they would take to try to increase efficiency of smaller, older data and so on. It's a follow through from strategy policy, which they have announced several times previously to try to be more assertive in forcing data center operators to become more efficient or even creating a kind of economic push for older data centers to shut down. It doesn't affect us at all. Joel, I mean, not negatively. I mean, maybe positively if there's some demand that shifts from that. Andin actual fact, I think we've played a very constructive role with the government in helping to establish this exercise, right, because it requires some IT and integration to create the platform for the monitoring of PUEs.
Yes, yes. Yes. I should add on something because GDS it's the most efficiency data center player in the market. I think this is not just a government asked us to do from our benefit point -- our self benefit point of view. This has improved the power efficiency is always our goal, our day-to-day job. So just like what Dan mentioned, it's not affect us.
Our next question comes from Alex Wang at Daiwa.
Management, could you hear me? Yes. Congratulations on the strong results. It's Wang Guohan from Daiwa Capital Markets. I will ask a question on behalf of John Choi. And the first question is regarding our customer diversification. As we just mentioned earlier, as we're trying to diversify our revenue mix towards non-Internet structures, do we have maybe a target for revenue mix from enterprise and the financial sector in the next 1 to 3 years? And also, we have observed your targets of MSR also middle single point decline in this year. And with our observation on new contracts occurred in the first quarter, is there any witness showed by the account to negotiate with new cloud contractor?
My second question is regarding our capacity injection plan as we see this year is more back-end loaded. So into the second half, what's our observation on the biggest downward risk for our capacity injection in the tech half because we can see that over 80% of our capacity actually acted in the second half.
Okay. Alex, we have sales targets, of course. But from a strategy and planning point of view, we don't set quotas for particular kind of customers or particular markets. So I mentioned that previously. I mean, I don't think that would be the right approach. I mean, we have tried to position to address demand from what we consider to be strategic, high-value customer base, and we want to be in the right place at the right time to capture that demand.
The comments I make about MSR reflect some internal assumptions about particularly about the pricing for new contracts, which itself is reflecting the location of that capacity, whether it's downtown, edge of town or remote.
And renewals also goes into that, although I'd say the renewals is a pretty small part of the change in MSR. We've commented quite a few times before that overall renewals are close to flat. We do have some older contracts, like 5- to 7-year old contracts, particularly for what -- for our first few data centers in each market. So it would be like the data center number 1, 2, 3, 4 in a particular market, where we had some, what we would call today, hyperscale customers. But at that time, their order size was a few hundred square meters, not -- and the total volume of their business with us has increased by 10 or more than 10x since then. Those customers, typically, they have downtown capacity and edge-of-town capacity. So when those contracts come up for renewal, and it's only a pretty small part of our total contract base, we've seen some very small decline of that there. But once you factor that into our overall MSI doesn't make much. It's not material difference. By far, the biggest difference is the location mix, the B-O-T projects, the edge-of-town and so on.
For the CapEx number, I mean, we -- if you look back, we added around 80,000 to 90,000 square meters of capacity and service in the last -- per annum in the last couple of years, and this year, we brought that down to around 60,000. So we've done that because of the slower move-in. But in order to do that, we have to consult with customers because there's also contractual delivery commitments that we have to adhere to.
If you're asking whether there was a risk from the point of view of being able to execute, I think that's relatively small. I mean, we're not affected by supply chain issues to any great extent in terms of our construction activity. We deal with suppliers in a very strategic way. We place what we call kind of bulk purchase orders. They produce a lot of equipment and hold it in inventory to be able to deliver to us at very short notice.
So we don't -- within Mainland China, we don't really experience material issues in terms of getting plant and equipment supplied. Maybe had some small effect on our construction in Hong Kong just because of the logistics from Mainland China to Hong Kong. But yes, so that's not really what's driving the 60,000 square meter number. That's more of a business decision to slow down the rate at which we bring capacity into service as a sort of ripple through of the slower moving.
Our next question comes from Sara Wang at UBS.
So I recall during our last earnings call that management shared that the outlook into next year should be better than this year. So given what happened so far, are we still holding the sales towards next year? And then what are our key assumptions for the outlook into next year?
Sara, I can't remember exactly what I said. I'm giving you the benefit of doubt if I said next year would be better. I don't know if you specifically meant a year because that's just an arbitrary start and an end date. But I mean, clearly, we're going through a cycle, right? Everyone knows that. And we will -- the cycle will bottom at some point, and there will be a recovery.
We have got many quarters of historic data on the ratio of move-in to backlog. And although there is quite a wide range, that has held for a long period of time. Currently, we see move-in to backlog ratio has fallen to a level that we have not seen historically. So it's well below historical levels. So I don't think that -- I don't think anything fundamental has changed or structurally has changed. I think that's cyclical and a reflection of some of the unprecedented factors which are currently affecting the market.
So once that -- once those go away and once we see the cycle turn, I don't see why the historical ratio of move-in to backlog should not return. And if that's the case, then we will see higher growth rates than we're seeing currently.
That's all the time we have for questions. I'd now like to 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 Group Investor Relations. Bye.
Thank you. This concludes this conference call. You may now all disconnect.