Hello, ladies and gentlemen. Thank you for standing by for GDS Holdings Limited Third Quarter 2021 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.
Thank you. Hello, everyone. Welcome to Q3 2021 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.gdsservices.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 prospectus as filed with the U.S. SEC. The company does not assume any obligation to update any forward-looking statements, except as required under applicable law. Please also 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 over the call to GDS Founder, Chairman, and CEO William. Please go ahead, William.
Thank you, Laura. Hello, everyone. This is William. Thank you for joining me on today's call. I'm pleased to report another quarter of solid results, with revenue and adjusted EBITDA up around 35%. During the third quarter, we made significant progress in key areas which underpin our future success. We sustained our sales momentum. We further diversified our customer base with another great hyperscale win. We enhanced our capacity pipeline with strategically important resource acquisitions. We took steps to secure our transition to renewables. We put in place further foundations for our existing platform expansion in Southeast Asia. In 3Q 2021, we booked 23,000 sq m of new organic commitments in Tier 1 markets. We remain well on track to hit our sales target of over 90,000 sq m organic booking for the full year.
As shown on Slide five, we won five hyperscale orders. Two orders were under an existing multi-year sales framework agreement with a top customer. Two were from one of China's largest internet platforms, which is a new hyperscale logo for us, demonstrating the strength of demand from internet. A further order was from a major Chinese bank, which highlights the growing potential of financial institutions. Our sustained sales is strong evidence that customers are not holding back on their medium-term and long-term business plans. For us to offer a complete solution to our customers, we must have continuous capacity supply in each Tier 1 market. Securing this supply means working with government policy. The government recognize the importance of data centers to the digital economy. The government also attach great importance to sustainability.
In its plans for the industry, the government has made clear that they want to see new data centers, which are large scale, highly efficient, smart, secure, and technologically advanced. They want data centers in Tier 1 markets used predominantly for low latency applications. They want a well-integrated edge, hub, and remote data center layer. Last but not least, they want data centers which are more green and lead the way in terms of renewable energy usage. GDS is well aligned with all of these government's objectives. Going forward, we expect new project approvals and energy quota will become even more difficult to obtain in Tier 1 markets. In some downtown locations, supply is already constrained. We therefore took the decision to accelerate our acquisition of qualified projects in key locations. Turning to Slide nine.
In Beijing, we closed the acquisition of Beijing's 17, 18, and 19, which brings us 10,700 sq m of developed and developable capacity. In addition, we recently signed an agreement for the acquisition of another 13,000 sq m of capacity, comprising four data centers on land owned by the project company, which are nearly complete and are not yet committed. We will market these data centers to financial and enterprise customers whose national headquarters are mostly located in Beijing. In Shenzhen, we are acquiring a data center with 3,600 sq m of capacity, which is nearly complete and already committed to our customers. In Guangdong, we are acquiring a portfolio of development projects at multiple locations.
The portfolio has a total developable net floor area of over 100,000 sq m, or IT power capacity over 250 MW. The projects are either at an early stage or held for future development. Guangdong Province has stated that new application will be extremely limited for the next few years. This acquisition enables us to materially step up our market presence. With the acquisition of Wuhan 1 and 2, we are entering a new market. Wuhan is an economic hub of central China, where many hyperscale cloud and internet companies set up their regional headquarters. The two data centers have a net floor area of 8,400 sq m, and the first phase is under construction. With the addition of this acquisition to our secured pipeline, we are well-placed in terms of supply in each Tier 1 market.
Turning to Slide 11. In a couple of weeks time, we will publish our inaugural ESG report and set out our sustainability targets for 2030. As is typical for data center companies, 99% of our carbon emissions come from electricity consumption. We estimate that for cloud and internet companies, around 60%-70% of their carbon emissions come from data center operations, including in-house and all sorts of data centers. It is therefore critically important to us, our customers, the government, and other stakeholders that we set proper targets and have a feasible strategy to reach carbon neutrality. Where are we today? In FY 2020, over 20% of our total electricity consumption was green. In the current year, we expect the percentage to be over 30%, and by 2025, we should exceed 50%.
We are increasing our green energy usage in three main ways. The first is by direct investment in renewable energy generation. We started in a small way by installing solar walls on some of our data center buildings. Our ultimate objective is to have large-scale integrated development. We are working with partners and the government to make this happen, but it will take time to realize. The second is by direct power purchase or DPP, which we are already doing pretty much to the maximum extent possible in Shanghai, Shenzhen, Chengdu, and Chongqing, as well as for some remote sites. DPP is currently constrained by the availability of renewable in Tier one markets, but supply will increase significantly over the next few years.
The third is by the purchase of renewable energy certificates or RECs, which in the near term is the most practical option. We recently signed a multi-year agreement with CGN New Energy, a state-owned independent power producer with a wind, solar, and hydro portfolio to purchase over 30,000 GW of renewable energy certificates. To put this into context, 30,000 GW is over 3.7 times our current annualized electricity consumption. We believe that this is one of the largest renewable deals done by any private industry user in China. In a further sign of progress, some of our data centers in Beijing were recently allocated tradable carbon credits as a result of our success in lowering PUE.
These GDS data centers are among the very first to be allowed to participate in Beijing's official emissions trading market. Now turning to Slide 12. Yesterday, we announced our second project commitment in Southeast Asia. We are in the process of acquiring a greenfield land for 28 MW data center development in Nongsa Digital Park, Batam, Indonesia. A special economic zone which is 35 km from Singapore. This is the new location for data center development, which you can see from the quote in our press announcement, has strong support from both Indonesia and Singapore governments. We expect to obtain renewable energy for this site, which will further enhance its marketability and competitive edge.
With this addition, we now have strategically located large scale project both in the north and the south of Singapore, with great potential for creating unique ecosystems and interconnectivity in and around Singapore hub. At the Nusajaya Tech Park in Johor, our land is next door to Telekom Malaysia's regional data center. We are pleased to have signed a strategic cooperation agreement with Telekom Malaysia, both for network connectivity and for use of available capacity in their facility, which will enable us to kick-start our presence before our own data center come online. Our regional strategy is driven by the requirements of our home market customers. To reinforce this point, we have recently signed a strategic cooperation agreement with a major Chinese cloud service provider to support its international expansion.
Under the agreement, we will be prioritized as their data center provider in the region. The strategic cooperation extends the mutual trust between us and the customer from China to Southeast Asia, providing a strong foundation for the success of our regional strategy. Now I will hand over to Dan for the financial and operating reviews. Thank you.
Thank you, William. Before going through the numbers, I'd like to provide an update on our partnership with GIC. A few years ago, we started to develop projects for our strategic customers in more remote locations using a BOT structure. We've done 15 to date, the status of which is shown on Slide 14. As you may recall, we established a partnership with GIC to finance these projects. Our original intention was with GIC, under which we will retain a 51% stake. With this higher level of ownership and additional fee income, we are able to deploy more capital while earning a reasonable return on investment. On the back of this new agreement, we've now signed the first sales and purchase contract with GIC for the transfer of a 49% stake in one of our existing BOT projects, Huilai 1.
We will work with GIC to transfer 49% stakes in several more BOT projects over the next few quarters. Starting on Slide 16, where we strip out the contribution from equipment sales and the effect of FX changes. In 3Q 2021, our service revenue grew by 10.6%. Underlying adjusted gross profit grew by 7.6%, and underlying adjusted EBITDA grew by 7.7% quarter-on-quarter. Our underlying adjusted EBITDA margin was 46.9%. Turning to slide 17. Service revenue growth is driven mainly by delivery of the committed backlog and closing of acquisitions. Net additional area utilized during 3Q 2021 was 18,678 sq m. The organic move-in, excluding acquisitions and BOT projects, was 16,037 sq m.
This is a step up from the levels seen in the first and second quarters. However, it was still a few thousand sq m less than our original expectations. We attribute this to a combination of server shortages, power uncertainties, and other macro factors which we believe are transitory. In 4Q 2021, we expect organic net adds to be at a similar level as in 3Q. Following the new agreement with GIC, we will now include area utilized of all BOT projects in the MFR calculation for 3Q and prior quarters. On a consistent basis, MFR per sq m increased by 1.3% quarter-over-quarter in 3Q 2021 to RMB 2,361 per sq m per month. This increase is within the normal range of quarterly fluctuations.
The inclusion of all BOT projects brings the MFR number down slightly, but the quarterly trend line is essentially the same. Turning to Slide 18, our underlying adjusted gross profit margin was 52.5% for 3Q 2021, a decrease of 1.5 percentage points quarter-over-quarter. Our underlying adjusted EBITDA margin was 46.9% for 3Q 2021, a decrease of 1.2 percentage points quarter-over-quarter. The margin decrease was mainly due to higher utility costs. In 3Q 2021, our utility cost was 28.8% of revenue, compared with 26.9% in the prior quarter, an increase of 1.9 percentage points.
Higher utility cost was partly a result of seasonally higher power consumption during the summer months, and partly a result of temporary additional costs for the use of backup power during a period when grid power supply was limited. In 3Q 2021, we spent RMB 19 million on backup power, where the unit cost of backup power is around three times that of grid power. Far, in 4Q 2021, grid power supply is almost back to normal. During October, the government introduced reforms to further liberalize the power market. These reforms require thermal power generators to sell all of their output through the wholesale markets and allows the electricity price to float between a regulated base tariff and a higher ceiling.
When fully implemented over the next year or so, these reforms will enable us to purchase substantially all of our thermal power through direct negotiations with generators. While this may eventually lead to a lower unit power cost, in the short term, we expect thermal power tariffs to go up because of elevated fuel costs. We'll start to see this during the current quarter. Accordingly, we estimate that our utility cost as a percentage of revenue will be similar in 4Q 2021 to the level seen in 3Q. Around half of our customer contracts in terms of billable revenue have pricing structures which allow us to pass on increased power costs.
Turning to Slide 20, our CapEx for 3Q 2021 was RMB 3.8 billion, consisting of RMB 3.2 billion for organic CapEx and RMB 575 million for acquisition consideration, mainly related to Beijing 17, 18, 19. As at the end of 3Q 2021, we had a liability of around RMB 1 billion on our balance sheet in respect of deferred and contingent consideration for past acquisitions. After nine months of the year, our total CapEx stood at RMB 11 billion, including RMB 3.6 billion for acquisitions and a further RMB 0.8 billion for landbank purchases in China, Hong Kong, and Malaysia. Based on our expectations for when the acquisitions which we announced today will close and the amount of total consideration which is payable up front, we expect CapEx in 4Q 2021 to be around RMB 5 billion.
Looking at our financing position on Slide 21, at the end of 3Q 2021, we had RMB 10.1 billion or $1.6 billion of cash on our balance sheet, and our net debt to last quarter annualized adjusted EBITDA ratio increased to 5.1 times. In the first nine months of 2021, we completed RMB 3.6 billion of debt refinancing, and we have a further RMB 1.3 billion to complete by year-end as part of our annual refinancing plan. You can see on slide 22 that we continue to refinance at a significantly lower all-in cost. We are beginning to see the benefit in our effective interest rate, which dropped to 5.5% during 3Q 2021, compared with 6.4% a year ago in 3Q 2020.
Turning to Slide 24, as a result of a lower than expected move-in rate in the second half of 2021 and recent increases in power costs, we now expect our total revenue and adjusted EBITDA for the full year of 2021 to be in the lower half of the originally provided guidance ranges. Accordingly, we are narrowing guidance to correspond to the lower half of the original ranges. The revised guidance is therefore RMB 7.7 billion to RMB 7.85 billion for total revenue and RMB 3.66 billion to RMB 3.73 billion for adjusted EBITDA. When we gave CapEx guidance at the beginning of the year, we explained that the number only included the acquisition which we knew about at that time, namely BJ 15 and that there could be more M&A not included in guidance.
During the year, we made the strategic decision to accelerate our resource acquisition, which effectively brings forward some future year CapEx. As a result, we now expect CapEx for the full year of 2021 to be around RMB 16 billion, compared with the originally provided guidance of around RMB 12 billion. Out of the RMB 16 billion, approximately 50% will be organic, which is consistent with our original estimate for organic CapEx. At the end of the current year, we expect to have around RMB 5 billion invested in capacity held for future development. We'd now like to pull up to open the call to questions. Operator?
As a reminder, ladies and gentlemen, hit star followed by one on your telephone keypads if you wish to ask a question. If you wish to cancel your request, you can press the pound or hash key. For the benefit of all participants on today's call, please limit your questions to two per participant. If you have more questions, please re-enter the queue. We have the first question. This is coming from the line of Jonathan Atkin from RBC Capital Markets. Please go ahead.
Thanks very much. So I was interested in the comments about the revised guidance, and you talked about kind of a slower pace of move-in. I wondered how does that affect your growth expectations for 2022? And then the higher utility costs that you mentioned due to fuel, how does that affect your margin expectations for next year?
Hi, Jon. We're not providing guidance today for 2022, and I, you know, I can only make some comments about, you know, the fourth quarter and you can form, I think, your own view. The move-in in the third quarter, I said, was a few thousand sq m lower than our original expectations. The fourth quarter is likely to be another few thousand sq m. In that respect, we'll start the next year at, you know, a few thousand sq m, you know, below where we originally expected to end this year. You know, I don't have a lot of bottom-up visibility far into the future, but I do believe that at some point we will see a rebound in move-in activity. You know, our customer contracts give customers the flexibility over how fast they move in.
It's part of the value of outsourcing, and it's what enables customers to make commitments in advance. We always see variance around our estimates because it's not up to us, it's up to the customers. We should not be surprised by this. Once again, we talk about the effect of higher utility costs. In the third quarter, you saw that utility cost as a percentage of revenue was increased by more than 1 percentage point. Partly it was seasonal, partly it was due to the higher cost of backup power. In the fourth quarter, the seasonality works the other way, and we no longer have a serious issue, so far, you know, in relation to power cuts.
We, you know, have potentially higher power tariffs as a result of, you know, liberalization of thermal power pricing and higher thermal power fuel costs. Therefore, we've assumed that in the fourth quarter, utility costs will be, you know, elevated once again, a similar percentage of revenue as it was in the third quarter. Once again, I think you have to form your own view about how long that will last. I'm sure it's transitory. We all know that. I can't predict the electricity price.
Hello, Atkin. This is William. I'll add on my view. I think that number one, we got the feeling is the market demand is still very strong and a lot of customer want to move in more quickly. As Dan mentioned, the server shipment is a problem in this year. It's impacted the move-in rate and also some carbon and the energy shortage that a lot of customer slow down their move-in. We hope this element will be solved ASAP. If that's the case, I think it will come to the normal, even a catch-up rate moving, right?
Again, it's too early to say estimate the next year's revenue guidance. One thing is that the orders still being maintained very strong.
Thank you for that perspective. I wanted to ask about the balance sheet. You touched on it in the prepared remarks, but can you maybe just kind of recap your thoughts on future financing needs given the higher CapEx that you're guiding to? You mentioned joint venture activities, but also kind of higher leverage. Putting that all together, how far into the future do you consider yourself funded, and what would be some of the financing options?
We're always fully funded in terms of capital for what we have committed to, what we've announced. We had $1.6 billion of cash on our balance sheet at the end of the third quarter this year, which is adequate for all the things that we are currently doing. If we need more capital, you know, we do have the option of increasing leverage at the HoldCo level. You know, we have a revolving credit facility there for $300 million, which we could potentially upsize. However, you know, as I mentioned before, we also think that it's a good strategy to bring in outside equity at the project level, like we're doing with GIC for the BOT projects.
You know, as you are all very well aware, there's a lot of private capital looking for data center exposure, and sometimes private capital comes at a lower cost than the public markets. We do want to have a hybrid approach, and we're looking actively at some possibilities for fund type structures, which would give us access to very deep pools of capital, potentially some v alue add from partners. Sorry, John, I just wanna add potentially some value add from partners.
Lastly from my side, Slide 39 gives the contract renewals and as a percentage of square meters in service. It's between 10% and 11% between now and year-end 2022. What is your thought or expectations around the pricing around those renewals?
Jonathan, most of the renewals are for, you know, downtown data centers. You know, the first thing I would note is that customers are holding on to practically all of their capacity when it comes up for renewal. The churn rate has been a fraction of a percent. You know, we've renewed some contracts flat. We've renewed some down to some degree. As I've explained before, it's not a mark to market exercise, like it is perhaps in the U.S. It's a strategic decision in the context of the overall customer relationship where we agree a renewal, you know, in the context of all the business that we're doing with that customer.
I don't think you can really look at it as a kind of completely market-oriented pricing.
John, I add a little bit, color.
Thank you.
John, I can add some color. Because in China, the different market demand and supply situation is different. In my view, the renewal is just like new sales. Maybe in some market, we will keep the flat price. Some market, maybe we'll increase price. Maybe some market will give our strategic customers some benefit. This is a mixed strategy. It depends on the different market supply and demand situation.
Thank you very much.
Thank you. Once again, ladies and gentlemen, as a reminder, for the benefit of all participants on today's call, please limit your questions to two questions per participant. If you have more questions, please re-enter the queue. We have the next question. This is coming from the line of Yang Liu from Morgan Stanley. Please go ahead.
Thanks for the opportunity. Two questions here. The first one is, in terms of the demand, I guess this is the time that GDS start to talk with customers in terms of their demand next year. Could you please share with us in terms of the outlook from the key customers, especially internet and the public cloud customers about the demand for new data center build out? Because we recently observed some of your customers, their CapEx are relatively low, so I'm not sure if the demand outlook for next year. The second question is, we noticed that GDS closed a lot of acquisitions in the past quarter.
Could you please update us in terms of the competition in those deals? Is it more intense than before? Is it more or less intense? How about the valuation trend, upward trend or downward trend versus deal closed several quarters ago? Thank you.
Okay, Yang, let me answer your first question. I think in that recent I went to China communicated a lot of our key customers. In general, I think they maintain their previous plan. But market structure a little bit shift. Some maybe some traditional hyperscale customer slow down a little bit. But in the same time, we saw a lot of the new hyperscale customer grow more faster than our expectation. So it's in general, I think the demand still maintain very strong.
The market a little bit of shift to internet and video company and e-commerce company. This is the fact based on my communication with our key customers. In general, the market is still very strong, and demand still very strong. In addition, in this year's practice, we saw a lot of financial institution. They increased their CapEx tremendously. It's very surprised to us. This year we also got a lot of the financial institution order, even bigger than before we expect. I think this trend also will still continue. Dan, maybe you can answer some acquisition.
There's competition for every acquisition. You know, we have a dedicated team. We try to identify the opportunities earlier. We monitor them. You know, there's usually a right time to move forward. You know, we try to leverage our expertise and track record in doing acquisitions to give you know, counterparties a high degree of confidence in dealing with GDS, which can be a factor, you know, not just a matter of price. If you look at the acquisitions which we've done so far this year and the ones that we've announced today, they are predominantly in and around Beijing and in Shenzhen and Guangdong.
Those are the places where, you know, there is already clearly a constraint on supply or, if not currently a constraint on supply, a constraint on new project approvals and energy quota allocation. I think everyone in the industry recognizes that having developable capacity in those places is going to be very advantageous because there is no organic option in the near future, or practically no organic option in the near future. You know, that means these kind of opportunities are highly sought after. There are a few other players in China, mostly private companies, you know, who have the capital and appetite to pursue them. Those are usually the players that we find ourselves in competition with.
As regards to multiples, I want to take this opportunity to clarify what we mean when we talk about acquisition multiples, because most of the recent acquisition have been of, you know, development projects at various stages, you know, all pre-operational. You know, we evaluate them in exactly the same way as we evaluate organic projects. For the purpose of communicating with investors, you know, we take the acquisition cost in terms of enterprise value, plus the cost to complete and divide it by the estimated stabilized EBITDA. That's the basis when we talk about multiples. I would say there's no discernible change in multiples, you know, as I defined it.
Thank you. That's very helpful.
We have the next question. This is coming from the line of Colby Synesael from Cowen and Company. Please go ahead.
Hi, this is Michael on for Colby. Two questions, if I may. Now first, you noted a few times that it's getting more difficult to get power approvals in certain tier one markets. You know, if you could unpack what's driving this, in turn, you know, what that means for the balance of acquisitions versus organic CapEx in those markets moving forward. Second, you know, you recently announced that you're entering, or you're doing more in Southeast Asia. Could you just give us some more color on what the next steps are as you look to expand further into the region? Thank you.
I think GDS strategy is we definitely will obtain the carbon quota by ourselves. This is the main strategy. At the same time, in the last couple of years, we also based on the market demand, short-term demand, we always flexibly to acquire a lot of resource because we want to capture our customers' demand. We always use the different strategy to fulfill the market demand. In general, we are still historically driven by our organic. In the future, we also will maintain this strategy.
Maybe I'll ask this second question about what are the steps that we're going through in terms of implementing our strategy in Southeast Asia? Excuse me. You know, our initial focus has been on the demand in and around Singapore. As you're aware, there's currently a moratorium on new data center project approvals in Singapore. We found from our home market customers that they have very substantial appetite for increased data center capacity in Singapore, but which cannot currently be satisfied. Therefore, you know, we are focused on coming up with a near shore solution which can address spillover demand from Singapore and act as a hub for serving the region.
You know, where we've got to so far is that we made an acquisition of land in a tech park in Johor, Malaysia, which is just 7 km from the Singapore border. Yesterday, we announced that we're acquiring land on Batam Island, on the other side of Singapore from Johor, which is 25 km from Singapore. Both of those sites are sufficiently large, in terms of what we've acquired and what we've got options over, you know, to do hyperscale developments, you know, hyperscale on a scale which, you know, cannot conceivably be done in Singapore, given the natural geographic constraints of Singapore. But it's not only hyperscale.
We are working to package these sites together with renewable power supply and multi diverse network connectivity into Singapore and between the sites to create a very well interconnected hub. We think that the combination of hyperscale, renewable energy, multi diverse network connectivity at a price point which is substantially below the price point of the Singapore data center market will be very marketable. Of course, as we're working on this, we're having back-to-back conversations all the time with our customers. You know, William mentioned today that we have signed a strategic cooperation agreement with one of our largest Chinese customers, which gives us priority in working with them for international expansion. It's a replica of the way that we've been working with them in China for the last five or six years.
You know, where we go next, we are looking for another site around Singapore. We do anticipate that there could be a relaxation of the moratorium in Singapore, although it may not be a very large. I'm speculating, but it would be very strategic to be able to establish a presence in Singapore. Then beyond that, you know, we would like to connect these sites into KL, where, you know, our Chinese customers already have cloud availability zones, and into Jakarta, where our customers also already have cloud availability zones. That would be, you know, the platform that we are, you know, aiming to get to at the kind of first base. Then, of course, that would be highly connected into Hong Kong and China.
That's as far as our vision extends for the time being. We've made very rapid progress. You know, every month, I would say, we get more excited about the scale of the opportunity that this represents. You know, in three, six, nine months' time, I expect there'll be, you know, plenty more significant developments to report.
I would like to say it's a super exciting opportunity for us because historically we refer a lot of our infra-based customer, so to say, demand to our shareholder STT. In the last five years, we saw the opportunity step by step increase. What we saw is the potential deal is much bigger than before. That's why we made a decision to jump into the market directly. It's the right timing, we think. It's like, sort of, let's say six years or seven years ago what is happening in China.
The scale is much bigger than six years and seven years ago. I think it's a huge opportunity for us because we have the very very huge very very strong, let's say, advantage in this area. Because number one, we bring all our potential installed-base customer. Their demand in South Asia, East Asia in next three years is very clear, very certain. This is number one. Number two, GDS builds not just one project. We try to develop a platform to solve our customers' pain points in this region, which make huge difference in the local player. In the meanwhile, we leverage our scale.
We can build very low cost data centers in this region, which we believe is a huge advantage because we have the scale, GDS has a scale advantage. GDS definitely will have super success in this region, which I believe in the next three or five years. This is where new incremental growth. This is not calculated in our previous five-year business plan. It's an addition.
Thank you. Our next question comes from the line of Frank Louthan from Raymond James. Please go ahead.
Great, thank you. Two quick questions. On the slower installations, any concerns there that you're losing any contracts or any competitors taking shares at any part of that? I wanna go back to the renewal question and ask that through a different way. You know, looking at that 10% of renewals, are there any significant customer or industry concentration between now and year-end that could add to the risk of the rent roll down that you might be facing? Thanks.
Sorry, Frank. I'm a different place from William. I will answer. Just to keep it brief, yeah, I think the move-in which we say is lower than what we originally expected, which means, you know, the view that we took, you know, nearly 12 months ago about what our customers would be doing in November, you know, 2021. You know, it's a few thousand sq m below. We don't like to talk about individual customers, but I suspect that the reasons for it being slightly slower are different from customer to customer. You mentioned competition. I would actually say that it actually may be the opposite of what you were thinking.
It's not a result of customers prioritizing moving in elsewhere. Generally, we are very fortunate and benefit from customers giving us priority in terms of move-in. They may have several data centers with capacity available to them, but you know, because of our relationship, we often benefit from customers giving us the favor of moving in to our facility ahead of moving into another service provider's facility. Again, for renewals, you know, we don't normally mention the churn rate because you know, I like to say it's statistically insignificant, but I think it was 0.2% last quarter. I think that tells you that there's no movement in the customer base.
You know, the data centers which are coming up for renewal, I say are downtown. Downtown means that they were, you know, amongst the first five, six, seven that we built in Shenzhen, Shanghai, Beijing. You know, the customers are holding on to that capacity. It's purely and simply a matter of a price negotiation, and William talked about some of the factors that we take into account when we conduct that negotiation.
Okay, great. Thank you very much.
We have the next question from Edison Lee, from Jefferies. Please go ahead.
Hi. Thank you for giving me the opportunity to ask the question. I will just stick to one for the time being. I think on the BOT projects that you are doing with GIC, the previous idea was to sell 90% stake to GIC, and then you collect a management fee and keep 10%. Right now, you want to go up to 51%. In terms of cash flow, does it mean that you are going to still collect money from GIC, but you will collect less because you are gonna own a higher equity stake in this project? How a change from 10% to 51% change your IRR prospect for these projects?
Edison, you're correct. You know, we have not changed the economic arrangement. You know, as a provider of management and operating services to these projects, GDS receives a fee which is, it so happens it's structured as a revenue share. We take a certain percentage of the revenue, and we cover our own costs. You know, we incur operating costs directly, and the revenue share that we take covers those costs and give us the margin. So we're charging the project companies the revenue share. Originally, we thought that we would be owning 10% of the project companies. Now we're owning 51%. So in that sense, yes, we're charging 51% of the management fees being charged to GDS and 49% to GIC.
Now previously, the way we were looking at this is, you know, we were trying to maximize management fees and minimize the capital that we invested. We thought that we could, you know, do these projects just as a kind of capital light approach. We look at the quantum of the management fees, and it's not insignificant, it's not enormous. And then we look at the opportunity to deploy capital, earn a return on that capital, and enhance it with management fees. We simply decided that it made more sense for us now and better, I think, for our shareholders for us to take the latter approach. It so happens that as we start to work for more different customers with these BOT structures, we did get asked by some customers to maintain a majority stake.
Although I think all the customers very much like the partnership approach and the involvement of GIC. It did come up a few times, so that was another factor for doing it.
Thank you. We have our next question coming from the line of Hongjie Li from CICC. Please go ahead.
Thanks management. My question is regarding the initiatives on green energy, because you mentioned in presentation that to do more direct investment in renewable energy, just like generation solar panels. How should we outlook the incremental CapEx and impact on the margin? Thank you.
Yes, we did mention that. I think that strategically our top objective is to be able to invest in some renewable power generation on an integrated basis with data center development. That would be highly innovative in and around tier one markets and, you know, it takes time to put together projects like that and time to develop them. That's an objective which I think will take several years to you know to realize it, to make it happen. In the meantime, it means that most of what we're doing in the area of renewable energy is OpEx.
You know, if you look at the utility cost, typically in tier one markets, it averages around 0.65 RMB per kWh for thermal power. Renewable energy certificates cost about 0.01 to 0.02 RMB on top of 0.65 RMB. Direct purchase of renewable power, which can only really happen in places where we are able to take delivery of renewable power as an end user. As a renewable power, you know, right now I would say the premium is typically around 0.05 RMB. RMB 0.05 on top of 0.65 RMB. That gives you some idea of what the cost is in terms of OpEx.
We haven't yet made any significant CapEx allocation to investment in renewable energy generation.
I will add on something color on that because our customers are all getting the pressure from the green data center or carbon neutral, right? They're facing more and more strong pressure. Based on this, we already talked to our customers. We believe the cost we will transfer to our customers.
Thank you. As there are no further questions at this moment, I'd like to turn the call back to the company for closing remarks. Please go ahead.
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. Thank you all. Bye.
This concludes this conference call. Thank you all for your participation. You may now disconnect your lines.