GDS Holdings Ltd (GDS) 2020 Q1 法說會逐字稿

Hello, ladies and gentlemen. Thank you for standing by for GDS Holdings Ltd first quarter 2020 earnings conference call. (Operator Instructions) Today's conference call is being recorded. I will now turn the call over to your host, Miss Laura Chen, Head of Investor Relations for the Company. Please go ahead, Laura.

Thank you. Hello everyone. Welcome to 1Q20 earnings conference call of GDS Holdings Ltd. The Company's results were issued via Newswire Services earlier today and are posted online. A summary presentation, which we'll refer to during this conference call, can be viewed and downloaded far our IR website at investors.gds-services.com.

Leading today's call is Mr. William Huang, GDS Founder, Chairman, 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 Miss 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 US 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 services is included in the Company's prospectus, as filed with the US 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, include 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, Chairman and CEO, Mr. William Huang. Please go ahead, William.

Thank you, Laura. Hello everyone. This is William. I'm here in Hong Kong, with Dan and Jamie. We're celebrating the first day of release from government quarantine. Thank you for joining us on to this call.

I'm pleased to report that our performance in the first quarter of this year was highly resilient, despite the challenging operating environment, we grew adjusted EBITDA by almost 50% year-over-year and achieved our highest-ever margin at over 46%. Sales momentum was strong, coming into the pandemic, and is even stronger now. In 1Q20, we achieved a record level of organic sales, with over 22,000 square meters, or 46 megawatts of net area.

In order to keep pace with sales, we scaled up our development program. We ended the quarter with over 110,000 square meters, or 260 megawatts of capacity under construction, which was nearly 70% pre-committed. We also added over 100,000 square meters of developable capacity to our pipeline.

Last but not least, we maintained our record of double zero, with zero infections and zero service incidents. We have been through a lot of turbulence since our IPO. Trade wars, pandemics, short attacks, even all three at the same time, but our business has not been impacted. We have kept our focus and demonstrated our execution capability. The outlook for our industry is better than ever.

Cloud is our main growth driver. Cloud aggregates demand from all types of customers, and all types of new technology. You may have seen reports that Alibaba plans to spend another RMB200 billion or nearing $30 billion, on infrastructure in the next few years. Other cloud service providers are also reporting accelerated adoptions.

Cloud in China is forecasted to grow at over 30% CAGR for the next five years. We estimated that this would require at least three times expansion of cloud data center capacity. The expansion will be balanced between Tier 1 market and the lower tier markets. This is the opportunity which we are talking to our customers about. Their procurement is consistent. They have high visibility in their business, and they are planning multiple years ahead.

As cloud service providers grows, they set up new availability zones, each of which is centered around an on-ramp. We focus on hosting these on-ramps when they're lower, and then on hosting the expansion capacity in the same location. The on-ramps also enable us to drive enterprise business. Increasingly, enterprises recognize that multi-cloud ecosystems as a key advantage.

We have had a lot of success with this strategy. As a result, we are well positioned for follow-on orders. To illustrate this point, our number 1 customer is now present in half of our data centers, and our number 2 customer is now present in one-third. This is what we mean when we say that we are the "Home of the Cloud" in China.

In 1Q20, we obtained three orders of over 10 megawatts each, the first time this has happened. We have many more prospects like this in our sales pipeline. At the beginning of the year, we set ourselves a sales target of 80,000 square meters of organic net add, a step up from what we did last year. 2Q20 is going well. By the middle of this year, I'm confident that we will be more than halfway towards our target.

Turning to slide 6, our data center development plan is entirely demand driven. We can see the evidence in our consistently high pre-commitment rate. It starts with joint resource planning with our customers. They share with us their specific requirements. We then go out to secure the right kind of land, buildings and power. Our objective is to keep our customers fulfilled on a continuous basis, both in Tier 1 markets and the lower tier markets.

In order to do this, we have evolved our resource strategy in a number of different ways. This is illustrated on slide 7. First, we have increased our ownership of property, and this gives us more options, more flexibility and more certainty. Across our whole portfolio, we are on track for a ratio of 60% owned, versus 40% leased.

Secondly, we have significantly increased the amount of capacity which are holding for future development. This is very strategic because it gives us the flexibility to respond to higher levels of demand. Our secured pipeline is enough to double our total capacity.

And thirdly we have expanded our market presence, in particular with our innovative asset-light approach to build-to-suit project in lower tier markets. This expands our addressable market, enhances our returns, and deepens our strategic customer relationships. It is still early days, but we see high volume growth potential with this product.

Turning to slide 9, in 1Q20, we closed the acquisition of a site in the Pujiang district of Shanghai. It is a good example of how we are solving the problem of resource supply. The site can support over 70,000 square meter net floor area when fully developed. Such a large amount of capacity on a single site within a Tier-1 market is unique. It is very advantageous for hyperscale customers as it enables them to land and expand in the same location. In the current quarter, we have begun the conversion of two existing buildings, which accounts for around 30% of the development. There's a lot of customer demand for this project.

A couple of years ago we started working with a strategic customer for build-to-suit data centers at one of their remote campuses. As of today, we have completed six build-to-suit(BTS) projects for them. Four of these projects are consolidated. There's a photo on slide 10. The remaining two are held pending sale to our joint venture partner, GIC, in the very near future. There's a photo on slide 11.

We have been pitching to other customers, and just recently, we were selected by a second major customer for their remote projects. We hope to partner with GIC again using the asset-lite model. Our customers needed to expand in low tier markets in sync with their expansion in Tier-1 markets. We are uniquely well positioned to address both opportunities.

With that, I will hand over to Dan for the financial and operation review.

Thank you, William. Starting on slide 14, where we strip out the contribution from equipment sales and the effect of FX changes. In 1Q20, our service revenue grew by 6.3%. Underlying adjusted NOI grew by 8.5%, and underlying adjusted EBITDA grew by 10.5% quarter over quarter. Our underlying adjusted EBITDA margin was 47.4%, the highest level we have seen. In 1Q20, we incurred a small FX translation loss. We target zero FX exposure. However, as we move cash into and out of China, it is unavoidable that we will recognize some FX gains and losses.

Turning to slide 15, service revenue growth is driven mainly by delivery of the committed backlog. Move-in during 1Q20 was 7,800 square meters, in line with our expectations under the difficult circumstances. Our MSR per square meter was down by 2.4% quarter over quarter. Most of this decline is mathematical because move-in was back-end loaded. In addition, there was a temporary reduction in some power tariffs that reduced revenue. If you look at MSR per square meter net of utility cost, the decrease was less than 1.0% quarter over quarter.

We recognize revenue as capacity is utilized in accordance with the terms of customer agreements and then we bill customers and collect cash with about a 70-day time lag on average. Slide 16 shows that our cash collection over the past three years is almost exactly equal to the revenue recognized. So far, we've not experienced any payment or deferral problems in the current situation. We have a very high-quality customer base. Apart from the cloud and large internet names which you know, the rest of our customer base is predominantly financial institutions and multinational corporations.

Slide 17 and 18 show the quarterly margin trends. A few things stand out. As a result of the reduction in some power tariffs, which I just mentioned, utility cost was more than 1 percentage point lower than normal. SG&A was also down, due to a combination of working at home, operating leverage and other government concessions. Some of this cost saving continued into April and May. Accordingly, we expect the underlying adjusted EBITDA margin to be in the 46% to 47% bracket in 2Q20.

Turning to slide 19, in 1Q20, CapEx was elevated, mainly due to the payment for the Pujiang site, which William highlighted. Now, organic CapEx was just over RMB1.0 billion. As at the quarter end, the cost to complete the entire 110,000 square meters of area under construction was RMB7.4 billion.

Activity on our construction sites is now back to normal. There are no supply chain issues to report. We have made minor revisions to the data center delivery schedule, which is shown on slide 20.

On slide 22, we show our cumulative unit development cost for self-developed area in service. We define "development cost" as gross PP&E, including assets held under finance leases. As you can see, our unit cost is now just over RMB78,000, or US$11,000 per square meter. This is for a data center portfolio which is almost entirely 2N redundancy. Over the past three years, we've increased the overall average power density to over 2 kilowatts per square meter.

The unit cost per megawatt is therefore around US$5.6 million, which we believe is exceptionally low by any benchmark. We've achieved this as a result of the scale of our construction activities, the efforts of our in-house team to standardize design and improve management of the supply chain, and by working in close partnership with our major customers.

We also show the NOI yield progression. For 2019, we achieved a return of 15.0% on area in service, with an average utilization rate of 69%. If we segregate out the stabilized data centers, as shown on slide 21, the return was in the high teens with an average utilization rate of over 90%. There are many moving parts to these calculations, but it all comes together in the returns, which you can see compare favorably with our global peers.

Turning to slide 23, we have two pending acquisitions. For Beijing 9, all the installed customers have entered into new contracts with us and we are managing the data center until the final CPs are satisfied. For Beijing 10, 11, 12, we expect to close by the end of the current quarter. When we announced the deal last December, two data centers were in service and one was under construction. The overall utilization rate was 50%. As of today, all three data centers are in service and the utilization rate has increased to over 70%.

Our M&A pipeline is strong. We are pursuing projects at the development stage and projects which are already ramping up. Data centers in China are attracting a lot of interest, particularly from financial buyers. This opens up new possibilities, including working with partners. We have many advantages in the M&A sphere and I'm confident that we will be able to conclude some more highly accretive deals.

Looking at our financing position, on slide 24, we ended 1Q20 with cash of RMB3.6 billion. We're now in the process of refinancing several of our recent property acquisitions in Shanghai and Hong Kong. Pro forma for this refinancing, our cash balance is back up over RMB5.0 billion. I mentioned earlier that the cost to complete our current area under construction is RMB7.4 billion. Assuming 40% of this is financed with equity, we will allocate RMB3.0 billion of our cash to capitalize these projects.

On the debt side, we have onshore RMB-denominated project finance in place, or in process, for most of the remaining requirement. We pioneered data center project financing in China and banks like the business. We've an excellent track record as a borrower and the banking market is highly supportive. Nonetheless, in order to give ourselves more flexibility over the timing of project finance, we are putting in place a large revolving credit facility at HoldCo level.

Our policy is to maintain a fully funded business plan. However, as is evident from our results, our growth is going to higher levels and the opportunity in front of us is expanding. From a strategic perspective, it is important for us to be in a position to respond to demand and keep our customers fulfilled. Therefore, we are always looking at ways of strengthening our capital base. We are well positioned in terms of access to capital, both from public and private sources, and will continue to evaluate all of our options.

Finishing on slide 25, our contract backlog now stands at 123,000 square meters, equivalent to 75% of our revenue-generating area. 75,000 square meters relates to data centers which are still under construction. Delivery of these contracts will commence over the next six quarters, as and when the data centers come into service. 47,000 square meters relates to data centers which are already in service. In 2Q20, we're expecting additional area utilized at over 10,000 square meters. Based on the first six weeks of 2Q, we are well on track to meet this target. It is a gradual recovery, which all being well, we expect to accelerate later in the year. Today we are confirming the full year guidance for revenue, Adjusted EBITDA, and CapEx, which we gave on our last earnings call seven weeks ago. With that I will end the formal part of our presentation and we would now like to open the call to questions.

(Operator Instructions) Jonathan Atkin, RBC Capital Markets.

Thank you. So I had a question about customer demand and then a question about the resource procurement and development. With respect to customer demand I just wondered if you can comment with respect to the overall market in China, any different trends that you are seeing with respect to Tier 1 demand versus more remote locations.

Then my question about development is just whether you have seen any changes with respect to challenges around resource procurement, the permitting environment for municipal projects, or edge of town projects, thank you.

Yes, Jonathan, I'll go first. Your first question about demand in Tier 1 markets and demand in lower tier markets. First of all, I think as we made clear, the overall demand picture has strengthened and this is the result of a number of factors. Demand is present in both Tier 1 markets and in lower tier markets and it grows in synch. The ratio for different customers differs but on an overall average for the hyperscale and large internet customers it's roughly 50/50.

So historically our business has been focused on Tier 1 markets. The last year through our partnership with GIC we found a way of opening up the opportunity to work with our customers in lower tier markets, and I think that is proving to be very strategic in terms of enhancing our customer relationships, as well as potentially being a very material opportunity in its own right.

I think, Jonathan, based on our bottom up conversation with our major customers I think the number they show us in the next three and five years is a very, very big number that we have seen. It's reasonable we think, because the Cloud their business and their visibility is very high so I think over time I think the China data center market is very exciting in a three or five years view, it's very exciting.

From a resource point of view, I think that GDS already took the action two or three years ago because we started seeking some larger land capacity and power capacity, because we believe this is a real fit for our customer future demand. So we are the first mover to focus on the edge of town to develop our resource. So we can say we are the first mover and we secured a lot of the land bank and power and it's not an issue for us to fulfil our customers' next three or five years business plan.

But, let's say, in urban town, it is still challenging for everybody in terms of the power quota policy, is still not very clear right now. Although a lot of the people talk about it the new infrastructure policy, if the new policy will release some more power. So far, we didn't see that but we are willing to see that. If that's the case, well, GDS can do more but we wait. On the other hand, we are quite certain we have locked up a lot of the edge of town resource which our customers like.

Great and then maybe just squeeze in a final question. On slide 39 you give the contract renewal schedule and if you could maybe just remind us at what point do you start having the discussions around renewing contracts that expire during 2020, 2021, and so forth, how far in advance are you having that dialogue?

Yes, Jon, just looking at the slide, yes, it's slide 38 in my copy about contract renewals. It shows 18,000 square meters of capacity or 6.3% of our total area committed comes up for renewal this year. If we drill down into the composition of that, there's no concentration. I mean the largest contracts that come up for renewal are single digit percentages of that 18,000. There are a few which pertain to cloud and internet customers.

Those are actually for our first generation data centers meaning Beijing 1 and Shenzhen 1, Kunshan 1, and so on. Those were the first deals which we did back in 2015, 2016, or maybe 2017. As I may have mentioned on a previous earnings call, at the time, these were unprecedentedly large deals, and now you look at it, it's 1000 square meters. It's not even worth calling out in terms of our quarterly growth. We believe that those customers will stay put. These are equivalent of edge data centers.

They put on ramps in some of these data centers, and I don't think they're going anywhere. The price negotiations I think we can say at least will be flat. As I also mentioned before, we have to look at this in the totality of our overall relationship with these customers, what we're doing with them elsewhere, what returns we're getting with them across the entire relationship. I don't think we can just isolate out one particular data center and treat it as if it's a completely standalone commercial negotiation.

Thank you very much.

Yang Liu, Morgan Stanley.

Thank you. Could management share your thoughts on the public REITs policy issued by Chinese financial regulator. What is GDS's plan on that, especially given the property ownership is increasing fast.

My second question is could you please update us on the customer move-in pace in the second quarter? You mentioned that there is no server supply issue anymore. Have you observed the meaningful recovery of move-in pace so far, and what is the current average month from a data center delivery to mature? Thank you.

Yes. Yang was referring to the pilot project announced by the National Development and Reform Commission and the Chinese Securities Regulatory Commission to promote the creation of a market for public traded REITs in China, backed by profitable infrastructure.

From our point of view, yes, it's interesting. We have a growing portfolio of stabilized data centers. An increasing proportion of them are wholly owned properties, and we should always consider whether it's beneficial for us to recycle capital out of such properties, while maintaining management and operational control for the long term.

The creation of a public traded REIT market might give us one option to do that, but there are many details still to be filled in in terms of how this pilot scheme is going to work. Meanwhile, we are already doing it in a way. We have mini-REITs through our partnership with GIC, and the largest financial institutions in China have approached us about the same kind of structure, and we can probably achieve the same result through privately negotiated deals.

Ten years ago, when we got into the data center business, we took a decision to structure each project where possible with a separate asset code for each data center, and then held in a separate entity, have the customer contracts and the operations. We always have in mind this would give us the flexibility to package assets or sets of assets in future, if it gave us better financing options. So we've always had this in mind. As I say, I think the REIT scheme, as it comes into fruition, may just give us one more option.

The second question about customer move-in pace, so it was just under 8,000 square meters in the first quarter, and right now, we're halfway through the second quarter, and I think we're looking at about 10,000 to 11,000 square meters in the second quarter.

It's already halfway to that level. It's going smoothly. In the third quarter and the fourth quarter, even in our original plan, we're expecting a step up, and to some degree it ties back to the completion of data centers, new data centers coming into service. We put a page in the presentation showing you how much capacity comes into service in each quarter in the remainder of this year and in the first half of next year.

And if you look at the backlog, you see at the end of the first quarter, we had about 47,000 square meters of backlog commitments, which relate to data centers already in service. Typically, we see that backlog be delivered at a rate of about 20% to 25% per quarter. So yes, if we hit 10,000 to 11,000 square meters in the second quarter, it will represent 20% to 25% of the backlog at the end of the first quarter.

And over many quarters, if you adjust for M&A and look purely at the organic business, you'll see that that kind of ratio holds consistently through. So I think if you use that methodology, it would lead you to forecast a high level of movement in the third and fourth quarter.

Okay, thank you.

Colby Synesael, Cowen.

Okay, thank you. Nice job on the name. On your prepared remarks, you mentioned that you guys have recently partnered with a second major partner, excuse me, for a remote site. I'm curious, was any of that reflected in the bookings or leasing the 22,000 square meters that you guys reported, and I'm just curious if that's ByteDance, because obviously we've put a lot of focus on them explicitly in terms of what they may be doing.

And then secondly, you mentioned that due to COVID-19, that the government had changed some things around power, which impacted both revenue and cost. I was wondering if you could be more explicit and actually talk about what the impact was, both for revenue and then the offsetting impact, which ultimately gets us to the NOI. Thank you.

Sure, Colby. I'll go first. To talk about the remote site business, we have one major customer that we've been working with, initially, on a consolidated basis, and then in the second iteration, through our partnership with GIC, and now we have activities at three different sites.

We completed four projects in one site, which was a consolidated projects, and we had a commitment to do seven projects through the joint venture. That's now increased to 12 projects, and with that customer, we're talking about a very significant increase in the volume of projects over the next couple of years.

And separate from that, a second major customer has now given us a purchase order for their remote projects, and the first one is going to be a trial. We have to go through the commercial negotiation and the design phase, but potentially, the volume from that second customer could be as high as from the first customer. It's not ByteDance. ByteDance is one of the few very large data center users in China, which is not yet our customer. We are aware of the demand that they have in Tier 1 markets, and we're trying very hard to fulfill that and to establish our relationship with them.

That's a high priority. William, do you want to add anything to that before I talk about power?

Yes, we are pretty focused on what we call it the three major cloud players, so as they mentioned that the next three or five years, it's a very big number in terms of both in the balance in the Tier 1 markets and remote. We tried to get their both market share as much as possible, so I think it's not including the ByteDance, because ByteDance, currently they are still deploy their servers in a remote area.

We hope we can do the deal like a remote area, like a JV with a GIC did it. But so far, we're still working on a couple of projects, which are at the Tier 1 markets. So our current number, our focus is not including any ByteDance expectations.

So, Colby, we are trying not to mix up the numbers for the joint venture projects, so in our 20-F and in our earnings presentation and so on, I'm keeping them quite separate. When you hear us talk about headline numbers like 22,000 square meters net add, that's all our core business. It completely excludes the joint ventures.

The joint ventures are disclosed on page 11. Those numbers are not included aggregated with our core KPIs anywhere. The other comment about power. Because of the pandemic situation, the government in China has had a number of schemes to provide some support to economic activity.

We've benefited in three principal ways. First of all, there is reduction in some power tariffs in some places, most of which we've passed on to customers. They did result in a slightly lower revenue and MSR, but also resulted in a lower cost of revenue. And then secondly and thirdly, there was a reduction in rent payable to our landlords, which is state-owned enterprises and also a reduction in contributions that we normally make to what they call the social fund.

Overall, if you look at our margin improvement, and I think you have to look at the metric that we call underlying adjusted EBITDA, because that strips out the effects of FX changes, our margin went from 45.6% in the fourth quarter of 2019 to 47.4% in the first quarter of 2020, so that's almost a 2 percentage point increase. I would say half of it was due to the government concessions.

The other half was due to some cost reduction from working at home, some operating leverage, and frankly, there was also some cost increases as we changed the operating procedures in our in our data centers. So yes, I think you can say that half the margin improvement was due to concessions and half was due to more factors under our control.

So those concessions were continued in April and they're continuing in May. We expect the margins in the second quarter to come out at a pretty similar level to first quarter.

Just one quick follow-up on that. So you said April and May is the stop, then if that goes away in June, or is that may continue, you just don't know?

Actually, the government policy is to continue, but I can't be certain. If you recall, our guidance for 2020, the midpoint of revenue and the midpoint of adjusted EBITDA implied about 46.3% adjusted EBITDA margin for the full year. So I think that still holds, and it's hard to normalize, right, because we'd have to normalize these concessions from the government.

We'd have to normalize the cost savings but also the cost - additional costs and also for the fact that we lost the revenue, right? So it's hard to normalize, but I think 46%-plus is most definitely still achievable for this full year.

Thank you.

(Operator Instructions). Gokul Hariharan, JPMorgan.

Hi, thanks for taking my question and congratulations on the continued good execution. First of all, just following up on the lower tier BTS projects, the total project size seems to be ramping up quite quickly. I think you talk about 48,000 square meters already and potentially more coming with a second customer.

Could we have a little bit of understanding about how the economics will work for GDS when you start collecting the management fee? What would be typically the IRR from that? Is that something, given the area is also increasing quite a bit, even though you only own roughly 10% of the equity -- how should we think about earnings contribution or EBITDA contribution from this kind of business, given it seems like this is declining quite rapidly in the last - I think, roughly in the year you've already added close to 50,000 square meters, that's my first question.

And second question, Dan, just to check, could you talk a little bit about how should we think about the MSR trend through the rest of the year, as we go into basically hold stable once some of these power subsidies kind of move away? And could you also maybe shed some light on path to profitability as your stable data center percentage continues to rise pretty quickly?

Okay, first on the lower tier data I want to make something clear, which is this is a completely different product from what we developed in Tier 1 markets. It's not just a different location; it's a different product. Typically, it's much lower redundancy and may deploy some technologies, like high voltage direct currents and so on, because it's much more cost oriented and less high visibility orientated.

So all of the metrics, which I hope you've become familiar with for our data centers and Tier 1 markets, they don't apply. You need a completely separate set of metrics to understand the unit cost and the revenue per square meter and the margin and so on. But as they're off balance sheet, I don't think it's worth going there, which would overcomplicate.

We do have a 10% equity involvement, it's literally only US$1 million or US$2 million or US$3 million per data center, so relatively speaking immaterial. Our approach was to maximize management fees. It's all sensitive, I'd like to tell you what we're going to make, but you can imagine it's commercially sensitive.

It approximates to let's say mid-single digits percentage of revenue of the data center as clear management fee profit. Once the data center comes into service, just like in Tier 1 markets, the customer has a move in period, which is factored into the return calculations and the pricing.

So the revenue ramps up, you can see from the projects we did in [Zhengbei] in ZB1, ZB2, ZB3, ZB4, that actually the projects ramped up quite quickly. As they ramp up, we will be earning that management fee and that kind of revenue share profit and that will come into our income statement as part of our service revenue. When it starts to appear, I will call it out, but for now it's not there and as I said, I don't want to overcomplicate.

I don't think we should talk in terms of an IRR, it would be off the scale, even compared with what I showed in terms of the NOI yields, simply because we have such a small equity investment, it's not that meaningful to talk about return on equity when essentially it's a managed product.

For MSR, I think the MSR in the second quarter will be quite similar to what it was in the first quarter and I don't expect much more decline over the remainder of the year. There will be some distortion, depending on when acquisitions close. That can also be a matter of timing, but also some acquisitions actually have lower MSR. But on an organic basis, we expect the MSR to be quite similar to where it was in the first quarter.

And the path of profitability I always protest that we have within our portfolio a large set of stabilized data centers. We don't segregate them out, they're part of the platform, but they are highly profitable, highly cash generative, not highly leveraged and so on.

If you look on a consolidated basis, in our 20-F we disclose the net profit, or loss as it so happens, onshore in China and the net profit or loss offshore in China. What you'll see is that the position in China is almost breakeven. So it's the finance costs which were booked at HoldCo level from the HoldCo level financing and the stock based compensation is also booked at HoldCo level, which is loss.

But for the business in China, the data center business, it's already almost breakeven. So I think if we go forward one year, even on a consolidated basis, it will be close to breakeven.

Okay, thanks, Dan.

Tina Hou, Goldman Sachs.

Hi, thank you, management, for your time in taking my question. Two questions from me. The first one is in terms of your electricity price for the data centers currently in service, what would the average price be? And also, in terms of the government-offered concession price due to COVID-19, what kind of discount did you get?

And the second question is I notice that you have an acquisition in terms of land in Chongqing. So wondering if we have any pre-committed customers for the site, what is the rationale of acquiring the land in there? Also I notice that for the previously Chengdu site the utilization rate is relatively lower versus the Company average, so how do we prevent this from happening again in the Chongqing site? Thank you.

The first question about utility cost, there are literally hundreds of power tariffs, so what we're talking about is an effective tariff rate. Your tariffs are structured with a fixed component, which is like a commitment charge, because there's a very large amount of generating capacity which is being reserved for the use of the data center and we have to pay that commitment charge whether or not we're using it.

That's why when a data center comes into service, it's part of the fixed cost of the data center and may lead to the data center being loss making until maybe it reaches 15%, 20% or higher utilization. The government concessions were actually mainly to reduce those fixed power commitment charges, so that's where we got the benefit. As I said in my prepared remarks, it was just in certain places.

Our effective power tariff, input tariff and frankly also the output tariff, because we pass it through, is about RMB0.65 to RM0.7 per kilowatt hour. We get the lowest rate which is applicable to large scale industrial users.

So your second question was about Chongqing.

I think Chongqing is mainly driven by one of our major customers. So this is we bought the land and we prepared to get an order in the second half of the year. I think it's highly achievable.

Tina, your second question was about the site we acquired in the Pujiang district of Shanghai. As we've been saying for many quarters, I think that is very familiar now and it's very challenging to obtain the right kind of real estate and power capacity in the urban areas. Nonetheless, it's absolutely clear, there has been no doubt, our customers still require it.

The edge of town capacity is a good substitute to enable them to scale up, but they still require the downtown data centers which are connected to the edge of town site. So we have to work very hard to generate new projects in the urban areas and we've had to try a number of different approaches, including buying building, doing acquisitions and so on.

In the case of the Pujiang site, this is an unusually large industrial complex, in a very good location, because it's close to where the submarine cable landing stations are in Shanghai. And we were able to have an agreement with the landlord that we would proceed with the acquisition of the property if and when, or as and when, we obtained the power capacity approval from government.

And then last year we did obtain that power capacity approval from the Shanghai municipal government and it pertains to this site, to this project, so that's why we went ahead. As William commented, hyperscale customers value a great deal the ability to expand on the same site, to put in an on-ramp and then to have visible expansion capacity on the same site. So having a site that's as large as this really has very unique marketing angles and we would regard it as a kind of crown jewel project.

Yes, let me add, I think in terms of Chengdu, I think in the whole portfolio, it looks like moving a little bit lower than our other major data centers. But what we can see is that Chengdu's data center move-in will catch up in the second half of the year.

Okay, thank you very much. And maybe just a quick follow-up, in terms of your CapEx guidance of RMB7.5 billion this year, I'm just wondering how much of that is land purchase and then the rest, I guess, will be the data center site construction.

Yes, the CapEx guidance was RMB7.5 billion. In round numbers RMB2.5 billion relates to acquisitions of data centers, as well as of property and land. So principally it consists of the payment for the Pujiang site and the payment for the Beijing 9 acquisition when it closes and the payment for the Beijing 10, 11, 12 acquisition when it closes.

So that accounts for most of that RMB2.5 billion number. The organic CapEx is around RMB5 billion, so it's riding at about RMB1 billion to RMB1.5 billion per quarter.

Frank Louthan, Raymond James.

Great, thank you. I wanted to touch on some new logos. What are you seeing from, in particular, your customer base of non-domestic Chinese companies? How is the sell-through going to them currently during the crisis?

And then secondly, what are you seeing from new logos in the quarter and quarter to date? How is that tracking? Thanks.

Yes, actually in the first quarter we won the very, very new logo. One is the largest coffee player in the world.

Not Luckin.

Not Luckin. So another very remarkable logo is one of the very critical as we call it foreign currency clearance house which is one of the branches of the Central Bank. That means Central Bank keeps giving GDS new orders in terms of the accepted China union pay, other clearance houses.

So we win a lot of the new meaningful new logos from it and in a big wave. So I think they've also contributed our new order booking in a significant way in the first quarter. We think this trend will maintain, because as I mentioned, now that a lot of retail customers are quite recognize the value of our data center multi-cloud access and the data center platform and the high quality of the operation. So we will attract more high quality retail customers, more and more in the next couple of quarters, which I believe.

Okay, great, thank you very much.

Due to the limited time of the call, I'd like now to turn the call back over to the Company for closing remarks.

Thank you, everyone, 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 and the Piacentre Group Investor Relations. Bye-bye.

This concludes the conference call. You may now disconnect your line, thank you.