GDS Holdings Ltd (GDS) 2020 Q4 法說會逐字稿

Hello ladies and gentlemen, thank you for standing by for GDS Holdings Limited Fourth Quarter and Full Year 2020 Earnings Conference Call. (Operator Instructions) 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 4Q and Full Year 2020 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 Chief Financial Officer, 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 safe harbor provisions of the US Private Securities Litigation Reform Act of 1995. Forward looking statements involve inherent risks and uncertainties. As such, the Company's result 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 USSEC. The Company does not assume any obligations 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, so the unaudited most directly comparable GAAP measures.

I will now turn the call over to GDS Founder, Chairman and CEO, William Huang. Please go ahead William.

Thank you Laura. Hello everyone, this is William. Thank you for joining us on today's call. GDS has been on an extraordinary journey for the past five years. The data center market in China has grown beyond imagination. As digitalization took off our growth trajectory has been unprecedented in the data center world.

We have become the clear market leader, reaching a scale which is multiple times bigger than our closest competitors. We have the best customer relationships, the most complete market presence, by far the largest development pipeline, the strongest balance sheet, the lowest cost of capital. Most important of all, an unmatched reputation which reflects many years of consistent delivery and high operating standards.

As we look forward from today, we see wave after wave of incremental demand. Driven by new technologies such as 5G, AI, cloud, and IoT, supported by highly favorable government policies. The market opportunities in front of us is inconceivable. While others are just waking up we are moving rapidly ahead to reinforce our market position by innovating with products and business models. Deepening our strategic customer relationships, adding substantially to our pipeline of scarce resource in Tier 1 markets. Enhancing our platform by entering new markets in China and overseas, seizing opportunities to further consolidate the market and the groundbreaking green initiatives.

We only just begun to reap the rewards of our past efforts. 2021 will mark our 20th anniversary. For me personally, GDS is still at an early stage of development. Over the next few years we will take the business to another level.

Despite the difficult operating environment last year, we made tremendous progress over the past year across every aspect of our business and have met or exceeded our expectations.

First of all, we beat our sales target adding over 136,000 square meters, or 271 megawatts of new customer commitments. We expanded our data center capacity in line with sales, adding nearly 140,000 square meters in service and under construction. We added significantly to our development pipeline, ending the year with 480,000 square meters secured for future development.

We stepped up our M&A activities, closing four deals with over 50,000 square meters of capacity. We grew revenue by 39.2%, and adjusted EBITDA by 47% year-over-year. Our adjusted EBITDA margin came out nearly 2.5 percentage points higher at 46.7%. We raised $2.4 billion of equity, and have successfully completed our Hong Kong IPO.

Turning to our sales achievement on slide 6. At the beginning of the year we targeted 80,000 of organic net add, plus 20,000 square meters from acquisitions pending closing. We overdelivered by a big margin, achieving 108,000 square meters of organic net added. Including 14,000 square meters from B-O-T project, and nearly 28,000 square meters from M&A.

Looking forward to 2021, we believe that the current level of organic booking is sustainable at around 90,000 square meters to 100,000 square meters of net add. Excluding B-O-T Data Centers. For the M&A part we already have over 19,000 square meters net added in progress from the Beijing 15 acquisition which is pending closing. We aim to close more deals this year.

Turning to slide 7. Every quarter a handful of hyperscale orders account for a large part of our sales. These 'must-win' are typically major deployments at our edge of town campuses. Hyperscale customers look to land and expand in key locations. For cloud service providers these locations are often configured as discrete availability zone which are critical to their IT architecture.

We are very strategic in targeting the first piece of business when customers deploy to a new area. We have succeeded many times in attracting hyperscale customers to establish an initial presence on our sites. It requires close collaboration with customers and right resource wherever they are going, and an undoubted reputations for delivery and operations. As a result of winning the first deployment, we have high visibility for a substantial amount of new business in 2021 and beyond as customers deploy additional phases on our sites.

Turning to slide 8. While we maintained great relationships with our top customers, our customer base now extended to almost all of the high-growth hyperscale names in China. We had exciting breakthrough last year, and we are working on more. The growth potential of some of our newer customers is extraordinary. We are highly focused on deepening these relationships, and there are significantly new business opportunities in the pipeline. We have established a high level of trust with our customers provided we have the right resource. We have an edge in winning new business. In the eyes of our customers GDS is not just an asset player, but a total solution provider. We have built our platform to mirror our customer's requirements and market presence. This fundamentally differentiates us from other players.

Turning to the Slide 12. It is clear that customers must continue to locate their mission critical, latency sensitive data center in Tier 1 markets. Customers target less than 5 milliseconds latency to core network nodes and between AZs, which imposes a distance limited of up to 100 kilometers around the urban center. It is also clear from years of government policy that the supply of suitable land and power in Tier 1 markets we have remained limited. Not just in the urban areas, but also in the surrounding edge of town locations.

We have recognized this years ago and have made huge efforts to secure sufficient resource to underpin our growth in all Tier 1 markets. As at the end of the 2020 we have secured 480,000 square meters of developable net floor area distributed across the key markets. Most of it is land which we have purchased from the local governments, together with allocation of power. This is a very valuable asset and another factor which sets GDS apart. We are not stopping at this current level. We have some big land deals in the pipeline and it will further strengthen our position.

Sustainability is an integral part of our resource strategy. The whole of China is grappling with this and it is not an easy problem to solve. We are working on a range of innovative solutions to source as much green power as we can. We are doing green power trading wherever possible and purchasing green certificates. We are also working with partners to evaluate co-investing in green power projects directly in the future. In 2020 over 20% of our total power consumption was green. In 2021 this ratio will go materially higher. We aim to publish our first ESG report later this year. We will set up targets and the road map which are realistic and achievable based on deep analysis.

In addition to the existing Tier 1 markets, we believe that some new tier markets will emerge in the next few years, particularly as a result of 5G and the need to push computing closer to the edge. Chongqing is an example. It has been on our radar for a while. We bought the land there early last year. We are now building our first data center on the site backed up with an anchor order in 1Q 2021. We are looking at another emerging Tier 1 market driven by customer demand. Over the next five years we could enter 10 new markets in China.

Turning to the Slide 14. Over the past few years an increasing number of data center projects have been started by independent developers who's objective is to sell. As a result, we see a window of opportunity to consolidate the market. We have an M&A track record like no other having done 10 deals in the past five years.

In 2020 we stepped up our efforts. We previously announced the SH19 and BJ15 acquisitions. Today we are announcing two new deals, both of them are data centers under construction but not yet committed by customers. They will give us highly marketable resource in their respective markets. We are paying a relatively small premium to organically build cost. We have a variety of M&A activities on our radar screen, some of which are sizeable.

Turning to Slide 15. A foundation of our strategy is to be a total solution provider to the leading Chinese customers wherever they have critical mass of demand. Our customers see a lot of value in working with a partner who understand their ecosystem. The same logic which takes us to the new markets in China leads us to look at expanding overseas.

Hong Kong is a start point outside mainland China. We currently have two major projects, the first of which is expected to come into service in 2022. We have recently secured an anchor [pass] order from Hong Kong 1 which we will announce in the next few months. The China cloud has big ambitions in South East Asia, both directly through their core platforms and indirectly through their strategic investments.

Take AliCloud as an example. They already have three AZs in Singapore, two in Malaysia and two in Indonesia. Singapore is a well-established hub for South East Asia and the global Tier 1 data center market. In recent years we believe that a large part of incremental demand in Singapore has come from our home market customers. For the time being the Singapore Government has suspended a data center project approvals, while new policies are developed around the land and power allocation.

It is uncertain whether Singapore given its resource constraints will choose to open the door wide for extensive hyperscale development. The adjacent markets in Malaysia and Indonesia are less developed than Singapore but have high growth potential. We believe that Chinese customer demand will be a critical success factor in these countries as well.

We have established a picture of demand from our core market customers. They have repeatedly requested us to establish a presence. We are actively pursuing opportunities with existing assets in Singapore as well as getting positions for when approvals restart. We have also entered into discussions with a number of potential local partners who have projects at various stages of development in Malaysia and Indonesia.

We believe that expansion into South East Asia is strategically important and that we can capture several hundred megawatts of new business over the next five years. We are moving ahead in a very carefully and deliberate way. We aim to announce several new commitments in South East Asia over the courses of this year.

To conclude my section, GDS is head and shoulders above everyone else in the China market. This is a matter of fact with what I told you today about the market opportunities in front of us. Our strategic positioning and our competitive advantages we believe that the gap is only going to get bigger. Now I will hand over to Dan for the financial and operating review.

Thank you William. Starting on Slide 18 where we strip out the contribution from equipment sales and the effect of FX changes. In 4Q20 our service revenue grew by 6.9%, underlying adjusted gross profit grew by 7.5% and underlying adjusted EBITDA grew by 6.2% quarter-on-quarter. Our underlying adjusted EBITDA margin was 46.8%.

Turning to Slide 19. Service revenue growth is driven mainly by delivery of the committed backlog and closing of acquisitions. Net additional area utilized during 4Q20 was 16,461 square meters, consistent with the previous two quarters. The first quarter of each year is usually slower due to Chinese New Year. Nonetheless, we expect movement in 1Q 2021 to be only a couple of thousand square meters down on the prior quarter's level.

Given the timing of capacity increases, as shown on Slide 23, we are forecasting that move-in over the course of 2021 will be heavily weighted to the second half. Monthly service revenue, MSR, declined 1.2% quarter-on-quarter in 4Q20 to RMB2,489 per square meter per month.

As shown on the next slide, MSR for the whole of FY20 was down 3.4% compared with FY19. In FY21 we expect a further low single digit decline. To some extent MSR is a reflection of average selling prices, however there are many other factors which affect MSR including the customer mix, data center location, redundancy level, development cost and contract structure. For example, the move-in flexibility and who pays for the power. Rather than talk about MSR on a standalone basis, I would prefer to focus on margins and returns.

Turning to slide 21, our underlying adjusted gross profit margin was 53.7% for 4Q20 and the same number for the full year, a near 1 percentage point improvement versus FY19. We calculate adjusted gross profit yield to enable investors to keep track of our returns in a simple way. It is a proxy for cash-on-cash returns. We divide adjusted gross profit for the year by the average gross amount invested, excluding assets or land under construction or held for future development. Gross amount invested includes the good will for all of our acquisitions.

For FY20 the adjusted gross profit yield was 13.3% which compares with 13.6% for the previous year. This was achieved with an average utilization rate of 70.1%, not materially different from FY19. Our commitment rate for area and service is 94.3%. When utilization rates catch up to the commitment level, yields can climb to the high teens. As you can see, we have sustained returns across our portfolio.

Turning to slide 22, our adjusted EBITDA margin had a slight dip in 4Q21 mainly due to one off expenses related to Hong Kong IPO events. For the full year, adjusted EBITDA margin was up 2.5 percentage points. Very approximately we estimate that a half percentage point improvement was due to net positive impact on costs of COVID. For FY21, we aim for about 1 percentage point of further margin expansion.

Turning to slide 24, our CapEx for FY20 was RMB 9.4 billion, a little bit less than what we guided due to timing of CapEx payments. RMB 9.4 billion includes RMB 6 billion of organic CapEx, RMB 1.5 billion of land and property purchases, and RMB1.4 billion of acquisition consideration. In 4Q20 we paid RMB 413 million of acquisition consideration, mainly related to the Beijing 9 and Shanghai 19 deals. We expect our CapEx for FY21 to be around RMB 12 billion including an estimated RMB 3.7 billion of consideration for the Beijing 15 acquisition and the 2 acquisitions announced today, assuming that they all close this year.

With respect to CapEx guidance, we can only include acquisitions to the extent that we have bottom-up knowledge. However, it is quite possible that there could be more M&A deals this year which are not reflected in our guidance. Part of the CapEx in FY20 and FY21 relates to organic projects which we built to suit for customers under build operate transfer or B-O-T contracts. There are different financial approaches to undertaking these projects. Previously we sought to minimize our equity investments and maximize management fees.

Now our preference is to invest more of our own capital, but to leverage it with lower cost debt. This approach gives us a reasonable risk adjusted return on equity. We therefore decided to keep on our balance sheets 2 B-O-T projects which we began constructing in 3Q20. Accordingly, we made some minor revisions to include these projects in our core KPIs, starting from 3Q20. We still have a further 9 B-O-T projects which we may partner with GIC that retaining a higher percentage of equity ownership. We will update you about these 9 projects in the next 1 or 2 quarters.

Looking at our financing position on slide 25, we had RMB16.3 billion of cash in our balance sheets and our net debt to EBITDA ratio is 2.2 times. Given our ongoing levels of organic CapEx and assuming the Beijing 15 acquisition closes shortly, this ratio will go back up to 4 to 5 times over the next few quarters.

Turning to slide 26. During FY20 we completed debt financings with a total facility amount of RMB16 billion, equivalent to $2.4 billion including both new project financing and refinancing of existing facilities. We made significant progress increasing the tenure of our facilities and lowering the interest rate margin. This is best illustrated by comparing the terms for 4 facilities which we refinanced in FY20, versus the original terms. The new facilities have tenures of 8 to 15 years, compared with 5 to 7 years for the previous facilities. Similarly, the new facilities have interest margins of negative 15 to plus 50 basis points compared with plus 120 to plus 270 basis points for the previous facilities. These improvements reflect the strength and track record of GDS as well as increased appetite for data center exposure from the banking sector.

In FY21 we anticipate doing about RMB15 billion equivalent to $2.3 billion of debt financing, including an aggressive refinancing plan. Refi involves some onetime costs which will impact our effective interest rate in the first 2 or 3 quarters of this year. We should then see a drop. Most of our RMB denominated debt is floating rate, priced relative to the over 5-year Loan Prime Rate or LPR. This reference rate currently stands at 4.65%. It is very stable, in fact it is hardly changed in the past five years.

What tends to happen when there is a tightening in the credit markets in China is that banks prioritize lending to favored areas, of which new infrastructure such as data centers is definitely one. We're therefore confident of achieving all of our debt financing goals this year.

Finishing on slide 28 with our guidance. For the full year of 2021, we expect total revenues to be between RMB7.7 billion and RMB8 billion, implying a year-on-year increase of between approximately 34.2% to 39.4%. An adjusted EBITDA to be between RMB 3.66 billion to RMB 3.8 billion, implying a year-on-year increase of between approximately 36.5% to 41.8%. In addition, we expect FY21 CapEx to be around RMB 12 billion. We'd now like to open the floor to questions. Operator.

Thank you. Ladies and gentlemen, we will now begin the question-and-answer session. (Operator Instructions). For the benefit of all participants on today's call, please do limit your questions. Our first question comes from Yang Liu at Morgan Stanley. Please go ahead.

Thanks for the opportunity to ask questions. I have one question on the competition. Could management update us in terms of the competitive dynamic in China market? Do you feel that the competition is getting more intensified compared with three or four months ago? Do we see the pricing return pressure or how does GDS do to defend the watershed in our key customers, when the peers are chasing them aggressively? Thank you.

Okay, good morning Yang Liu. First of all, I think the answer is in general I think nothing changed, in terms of compared with the last couple of years. We didn't see GDS position in the market will be changed. So, our position is very solid in terms of our platform, we are the only platform player in the market still. There is no platform player in the market as we mentioned in the last couple of the earnings call, this has not changed. The competitor or the regional player or private player. So the platform is very valuable for all our key customer, there's number one.

Number two, I think our position in our customer is very different than all of our follower. Our position is very solid and in our customer's eyes, GDS is very, very reliable platform, so this will empower their business in the past, even in the future. We're the only reliable platform in our key customer's eyes. So, what I can tell is of course a lot of this, because the market is booming, a lot of the new player in the market, even a lot of the established data center players. The competition between them looks like a more stronger than before, but not reflected to us.

So, in terms of the price and the return, I would like to say the price is very difficult to talk about the price, as we mentioned last couple of times. We have pursued the project return. I would like to say our future project return where we maintain what we talk about before. So, in general will not impact our project return.

Thank you. If I may have another question, I saw that we have a new financial metric underlying adjusted GP. Should I assume that it is almost the same with NOI we have been using for the past years?

Yes, the short answer is it is exactly the same, it was a different name and maybe it's a little bit simpler because gross profit is a standard accounting line item and adjusted growth profit just needs to be reconciled to gross profit. So presentationally I think it's a little simpler, but the number itself is exactly the same, yes.

Thank you. Our next question comes from Jonathan Atkin at RBC Capital Markets. Please go ahead.

Thanks very much. So, I'm interested in the slide that talks about your largest orders in Q4 and two of those contracts commenced already essentially, so that was a very fast turnaround between contract signing and contract commencement, and then the other two were fairly far outdate, appear to be built to suits. But then you talk a little bit about the nature of the customer demand that you're seeing? Is it similarly bound between what would appear to be immediate needs and then needs that are kind of put out?

My second question relates to CapEx and I wondered if there is any trend that you are seeing in your business around the build cost, construction cost per megawatt of IT capacity. Is that trended differently recently or is there any chance that that could decline over time? Thank you.

Yes, John, I answered first question and we highlighted four hyperscale deals. The first one, we acquired a data center in Shanghai as one of those projects that was undertaken by an independent developer and after we acquired it, we were able to move forward with the customer contracts. So that is why it there appears to be a short time lag.

The second one, which also appears short, is Beijing 7 that in this case, it is an existing data center. It was already, roughly speaking, half committed to customers and this was not the first order for the data center. This was the order that took up most of the remaining capacity. So once again, that is why there seems to be a relatively short time period because it was an order for an existing data center.

Now, the other two hyperscale orders in Langfang and Huizhou, yes, they both involve, in one case, greenfield assets and in the other case it is a conversion but it is an early pre-commitment right at the very beginning of the project. Sorry, John, can you repeat the second question?

(Technical difficulty) in China, I wondered if you have seen that, what the reasons might be or if you're not seeing that at all?

I am sorry, on my line I couldn't hear it clearly.

Construction you go from megawatt of construct

Yes.

So construction cost per megawatt of the IT capacity. Has that trended differently?

Yes, sorry, I couldn't hear clearly. The trend in construction costs has been that we are able to keep achieving cost reductions for a variety of ways. Some of it is simply scale and procurement. Some of it is to supply chain management. Some of it is through the way in which we approach the construction in terms of the phasing and the modularity and off-site prefab and so on.

Some of it, frankly, is by setting up projects which have the most optimal cost due to location, due to the proximity of power infrastructure. Due to the ability to leverage existing substations and so on. The gains year on year are small percentage gains but they are continuous and looking forward, we continue to see opportunities to continue this. So we expect to actually for quite a few years to come, to be able to lower our unit development cost by small increments year-on-year.

Thank you.

(Operator Instructions). Our next question comes from Colby Synesael at Cowen and Company. Please, go ahead.

Hi, this is Michael on for Colby. Two questions, if I may. First, it appears you re-stated your area under construction and a few other KPIs in the third quarter to include the B-O-T data centers. Just to be clear, this is because you intend to retain 100% ownership of those assets? If so, what drove the change versus your prior view?

Then my second question is, based on the conversations you're having in the markets outside of China to attempt to expand, which market would you expect to have a footprint in first and what would you consider to be a reasonable timeline for having a datacenter up and running there? Thank you.

Okay Michael, on the first part of your question, we did previously disclose the existence of these projects. We had a category we called Manage BTS Data Centers and in that category, we grouped them for projects relating to customer 1, projects relating to customer 2, projects relating to customer 3.

There was one project relating to customer 2 and one project relating to customer 3, so they were just single data centers on a site and we decided, because of the situation that it would be more optimal for us to undertake those ourselves in the usual way.

Therefore, we moved them from the category of what we called managed BTS Data Centers and then include them in our KPIs just as we would with any other self-developed data center but we do identify in the detailed breakdown in the appendix to our presentation, which are the B-O-T projects so that it's clear for all to see.

Now, more generally, with regard to the B-O-T projects, it is becoming quite popular, we're seeing more of the large cloud and internet companies in China experimenting with this approach for development on their campuses in remote areas. It's a situation where they clearly have the option to self-build because typically it's their campus and there is no barrier in terms of land and power but they still see advantage in outsourcing but the terms of outsourcing are quite different.

From our perspective it is not, yes, the core strategic focus but it is an adjacent part of our business and I think we would be flexible about how we approach it. In the past, we have done some on balance sheet and we put about 80% leverage on the projects and the return on equity to us was quite acceptable and I think going forward, we will do some that way. We will do some where we have a majority and maybe we will do some where we have a minority or even just purely a management fee.

I think the market will call for a wider variety of business models and certainly as regards these projects outside of the core markets, we will be adopting a variety of different approaches.

William, do you want to answer the second part of the question?

Yes.

Yes.

Okay.

Good on you.

Yes, I think the overseas strategy as I just mentioned, the first point is the start point is Hong Kong. We were really developing Hong Kong to broaden the development. So the next step is obviously it is South East Asia.

In South Asia, the top three countries in our radar screen is Singapore, Malaysia and Indonesia. These three markets is, in our view, in the next five years, the demand from China will be a few hundred megawatts. So we are trying to catch up with this wave but potentially, I think South East Asia is a high potential market.

We go to, our current strategy is, we have to set up our presence here but it is for the strategic reason but if we look at the current South East Asia market, it very similar like eight years and 10 years ago, the market in China so we are willing to step in and build our data center position, our position there and to catch up with the future high growth. This is our current goal.

So maybe we will open the first project will be in Singapore, maybe in Malaysia or Indonesia. So we will give you the clear answer in the near future.

Thank you. Our next question comes from Tina Hou at Goldman Sachs. Please, go ahead.

Hi, thank you very much for your time Management. So yes, my question is also related to the South East Asia expansion strategy. Wondering what kind of differences in terms of revenue as well as return in these potential South East Asia projects and how that may impact our P&L in the future and when do you see the first project to start contributing revenue to our P&L as well? Thank you.

I think it's too early to talk about data revenue because what we in the next five years, I don't think it will impact a lot in our revenue or because our current order guidance for the development is still, we are still in China because China is still a big market.

So as I just mentioned, South East Asia in our view is strategic but we have not calculated any number in that P&L right now. I think maybe when the projects come in, we will talk about that.

Our next question comes from Gokul Hariharan at JP Morgan. Please, go ahead.

Thank you. So could you talk a little bit about what are we seeing from national government policy in China regarding energy efficiency, pollution control et cetera given some of the bigger commitments on climate change?

Could you, I mean, I know that GDS would be publishing the ESG report this year but could we start talking a little bit about what are we seeing from a PUE perspective? Especially for some of the newer Asia fund projects and what GDS has been able to do along with customers on this front?

The second question is on M&A. It seems like M&A is now becoming a much bigger part of pipeline as evidenced by last year. If you think about the next two to three years, should we assume that M&A becomes probably much bigger part than even last year? Last year, I think it was about 20% of the total pipeline build. Should we think that M&A would be a much bigger part of the pipeline if we think about the next two to three years?

William, do you want to address the question about government policy?

I think the ESG is a very hot topic right now in greater China, including Hong Kong or mainland China. I think this definitely is a trend when the government encourages the green power and also, we call it (spoken in a foreign language).

Carbon neutral.

Carbon neutral, right. GDS already start to prepare for this, for a couple of years. As I just mentioned, 20% of our data centers are already green, green energy.

So, government's policy, from the policy point of view, I think they are in general still under development right now. We didn't see any official enforcement policy right now, but of course, the immediate impact is, as I mentioned, in the carbon quota in the Tier 1 market.

Especially also in some edge of town of the Tier 1 market will be getting, it's easy to see if it will control more tight in the future. This is what we can see, what's the impact, yes.

To answer the question about M&A, the main driver of our business for the foreseeable future is going to be organic development in Tier 1 markets, but having said that, we clearly see an opportunity right now over the next few years to consolidate the market.

There's been a lot of new entrants, a lot of development by independent developers, and that's what creates the pipeline and the opportunity to consolidate. We believe that we have significant competitive advantage in terms of M&A, in terms of the capability and experience of our team, the methodology that we developed, the deals we've done since 2016, and our financing capacity, in the ability to conduct technical due diligence.

We find that sellers, when you engage with sellers, always top of their mind is if they engage with a potential buyer, they want to know is that buyer going to get to the finishing line.

I think buyers, sorry, sellers have a lot of confidence when they engage with GDS, because of our track record. People ask about data center M&A in China, but I think most of it is being done by us. I'm not aware if very many M&A deals are being done by others. So, we are the major force in data center M&A in China.

The question about how much of this could there be, and frankly, we don't have a quota. We don't look at it like that. There's really two different types of deals.

One type, which I call a kind of flow deal, is the 5000 to 10,000 square meter data center typically that single data center on its site. It could be purely capacity, like the two deals we announced today at Shenzhen, eight in Tianjin 1, or it could be capacity with some customer commitment either there or coming with our acquisition, like in the case of Shanghai 19.

There's a steady pipeline of those kind of deals and the acquisition multiples have not increased, and we can still do these deals paying a relatively small premium to organic cost. We count in terms of our metrics based on when deals close, so if we do two of those it comes to maybe 15,000 square meters, we do three, it comes to something over 20,000 square meters.

The other kind of deal which is harder to predict is the more sizeable ones, let's say 20,000 square meters and upwards. We've done two or three, depending on how you categorize. There was the Beijing 10, 11, 12 deal, that was 20,000 square meters. There was the Beijing 15 deal, that was 20,000 square meters.

If you could call it M&A, there's the deal where we're partnering with CITIC which, by the way, is now upgraded to 28,000 square meters. These opportunities are scarce and hard to predict when they're going to come onto the market.

If they are at a more advanced stage, they have customer commitments, then there's likely to be more competition for those kinds of opportunities, and that's where multiples have probably gone up, but still I think reasonable and justifiable.

We have some opportunities like that in our pipeline and there can be quite a bit swing factor if they crystallize, but I don't want to set any expectation on that. That's where that could make quite a big difference. That's where we could go from having 20,000 or 30,000 square meters of acquisition net add a year to having something quite a bit more.

Our next question comes from Rob Palmisano at Raymond James. Please go ahead.

Hi, guys. It's Rob on for Frank. I just wanted to follow up, you sort of spoke to this earlier. Can you talk about likely sources of capital for next year, and also can you speak to your strategic relationship with CyrusOne going forward after they just recently monetized their investment in you guys?

Rob, we always try to be, or we have been fully financed, fully capitalized for our business plan and I think we're quite open about sharing what our business plan is.

I think based on the comments William made about the expected level of organic net add at 90,000 to 100,000 square meters, the fact we've got one 20,000 square meter acquisition already in the pipeline and potentially closing shortly, and hopefully other acquisitions that will close this year, that's the business plan which I think everyone can see, and we're well capitalized for that, at least for the next couple of years.

Look, we haven't factored in regionalization, we haven't factored in M&A beyond the flow deals, as I just described to you. We haven't factored in a higher level of organic growth, and to some degree, we haven't factored in the potentially bringing forward of CapEx if we see opportunities to acquire land or buildings which would be opportunities that are onetime opportunities that you have to grab when they're there, otherwise they're gone forever.

If any of that materializes, it's upside. It's upside to our business base case and believe it will be positive for our shareholders. If a need for capital arises, it should be seen as something, it will be because of greater success. William, do you want to comment about the relationship with CyrusOne?

Okay. I think in terms of the sales cooperation, we still maintain the sales cooperation relationship. Because historically, they refer some deal and we follow the deal, and they have us cut a couple of deals. But since now Chinese customers go to US, it's a slowdown, so they get a limited benefit from us.

I think in a way our relationship is still there, but the GDS current, our international team work very closely with the different partners, for example, last year we got a lot of deal from US, US-based customer, let's say, in China, independently. We are not relying on them, of course, I think they, but in some ways, we work together with some deals still.

Our next question comes from James Wong at UBS. Please go ahead.

Good morning, management. This is James Wong from UBS. I've just got two questions. The first one is a follow-up on the ESG. I'm just wondering, since Chinese Government has announced the carbon neutrality goal, whether your conversations with your largest cloud customers have changed.

Related to that, how does your current land held for development, future development, how much of that is for renewable energy and whether you'll be able to increase the percentage of renewable energy usage for your pipeline of projects. That's the first question.

The second question is on you expanding into, William mentioned 10 different regions in China. If I remember, for example, I also saw that you've recently got into Tianjin and Chongqing.

If I remember correctly, in the past, for example in the west the utilization rate increase has been rather slow, so I'm just wondering what you're seeing in terms of demand in these new regions, whether the situation has changed versus the past. Thank you.

Dan, answer this first question?

Yes. Sorry, I'm in a different place from William. I can't hear everything clearly. James, what I heard was you asking about the area held for future development. What proportion of the power is held with that?

Yes, so I'm asking about, yes, sorry. On the ESG front, conversations we align with our customers, where that is changed, where they're taking more focus on renewable energy usage, and also on your land held for future development, how much of it is renewable energy and whether that percentage can increase. Thank you.

Yes, sure. I think the first part of the question is everyone in China take more notice, start from the central government downwards. Our largest customers look to us and expect us to help them solve the problem, the challenge, of how to reduce their carbon footprint.

Probably the biggest component of it is where does the power come from, meaning what fuel is used in the power generation. It doesn't necessarily go with individual projects. For example, in a region, like within Shanghai, for example, we can purchase green power through the power trading market. That's one of the ways in which we reach the level of green power that we mentioned, over 20% of our total consumption. That's not location specific.

On the other hand, there are some locations where specifically there is green power because for example, the location may be on the same grid as where there is large amount of wind power generation, even though there could be quite a few hundred kilometers apart, it could just be that they are on the same power grid. We will come out with targets.

I think rather than talk specifically about which project has what, I think the way to look at it is in the aggregate. We will have to be innovative and take a number of different approaches, which we're working on all now at the same time. Whatever targets we set will be very serious ones. We take this very seriously; it's absolutely critical to our business. William, do you want to answer the second part of James's question about...

Yes, the second. Yes, sure. First of all, I think in China, the Tier 1 market is very obviously at Shanghai, Beijing, and Guangzhou. The western side, if you look at the last three years, actually they grow, the data centers demand in Chengdu and Chongqing, in last three years grows very fast.

So, this is the trend. So, given the current utilization, what we can tell is that the growth trend is very obvious, so we are very confident to increase the utilization in our current data center in western parts of China.

But another, in our radar screen, there's another 10 cities like Chongqing. We pay more attention on that. This is a new chance. In the last couple of years, what we can tell is a lot of our customers, our current customers, cloud player and a lot of the internet giants, they start to deploy their server in this city.

Look, we are very confident this ten new markets , will drive another growth in the future. I think as I mentioned, 5G in China is very advanced than other countries.

So 5G, what the impact is, when the 5G have completed deployment, it will produce more data from those cities, because those cities are economic centers in all the province and the population is, most of them is around more than 8 million or even 10 million.

So, I think there is potential market, now is just the start. We keep them like a new emerging market, but in order to make the investment more efficient, we will, we always will follow up our customer demand to get into this market. Yes, this is our view.

Thank you. Ladies and gentlemen, due to time constraints we have no further time for questions. I will hand over to GDS for closing remarks.

Thanks, everyone, 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. Thanks, all. Bye, next time.

Thank you. This concludes our conference call. (Operator Instructions)