NEXTDC Limited (NXT.AX) Earnings Call Transcript & Summary

Thank you for standing by, and welcome to the NEXTDC FY '21 Results Announcement Conference Call. [Operator Instructions] I would now like to hand the conference over to Craig Scroggie, Chief Executive Officer. Please go ahead.

Thanks, Travis, and good morning, ladies and gentlemen, and welcome to the NEXTDC results presentation for FY '21. I'm joined today by our CFO, Oskar Tomaszewski. Beginning on Slide 2, we're pleased to present another set of record results. Revenue of $246 million, underlying EBITDA of $134.5 million, exceeding our upgraded guidance of $130 million to $133 million. Contracted utilization now stands at 75.5 megawatts. Our network connections continue to expand, growing to more than 14,700, up 13% in the last 12 months. The ecosystem continues to evolve with 1,540 (sic) [ 1,547 ] customers, 730 partners and 70 connectivity service providers. On Slide 3, our strong performance in FY '21 is highlighted by robust key operating metrics. Revenue from data center services increased 23% to $246 million. Contracted utilization increased 8% to 75.5 megawatts and interconnection accounted for 7.7% of recurring revenue. The results further demonstrate the company's operating leverage with underlying EBITDA increasing 29% to $135 million. Operating cash flow grew 148% to $133 million and billing utilization up 24%. We remain well capitalized to support our growth plans. Total liquidity over $1.7 billion, inclusive of undrawn debt of over $1 billion, which achieved financial close in December '20. Our balance sheet is underpinned by a total of $2.6 billion in assets. Our data center fleet continues to evolve at a rapid pace. The fit-out of our S2 and M2 facilities in the gateway regions of Sydney and Melbourne continues with 17 megawatts of capacity built this year. S3 development is on schedule for practical completion in the second half of '22 with building construction well progressed. M3's planned capacity has been increased to 150 megawatts, thanks to recent further land acquisitions. Groundworks are well progressed and building construction now underway, on target for practical completion in the first half of '23. And finally, we announced our first hyperscale campus in Western Sydney, adding 300 megawatts of development pipeline as the land is progressively settled between 2H '24 and 1H '25. I'll now hand over to Oskar.

Thank you, Craig. Let's now turn to Slide 5, a summary of our profit and loss for the year. The statutory results reflect data center services revenue of $246.1 million, an increase of 23% on the same period last year and net loss after tax of $20.7 million. Our nonstatutory highlights include, as mentioned by Craig, an underlying EBITDA of $134.5 million, which is an increase of 29% on the same period last year; direct costs of $42.9 million, which rose in line with customer power consumption, offset by falling energy costs as well as improved efficiency; facility costs, which increased to $24.7 million, as we ramped up operations across our facilities, including the full year of operations at P2 as well as increased property holding costs and the addition of the M3 site; and finally, corporate costs increased to $43.4 million, reflecting the investment NEXTDC is making into the ongoing growth of our business, as well as the market-wide increase in insurance costs, particularly for D&O cover. On to Slide 6. Revenue generated from racks, suites, cross-connects and other recurring sources, accounted for 96% of total revenue with project income representing a lower proportion of revenue each year. The underlying EBITDA performance highlights NEXTDC's operating leverage as demonstrated by the 29% growth in earnings relative to 23% growth in data center services revenue over the past 12 months. This trend is further demonstrated by the longer-term earnings and revenue CAGR figures of 37% and 22%, respectively. Slide 7 sets out our revenue per unit metrics. Annualized revenue metrics continued to perform strongly throughout FY '21, moving in line with lower power costs, which are largely passed through to customers. Revenues from larger ecosystem enhancing customer deployments increase over time, due to higher usage of contracted power capacity, increased demand for interconnection as well as the use of ancillary services over time. Slide 8 summarizes our balance sheet position and cash flows. At 30 June, NEXTDC owned property with a carrying value of $997 million as well as plant and equipment with a carrying value of $785 million. Our net assets stood at $1.7 billion. And finally, we remain well capitalized to continue our growth trajectory with total liquidity comprising cash and undrawn debt facilities of just over $1.7 billion. I'll now hand you back across to Craig to go through our business performance and outlook for the 2021 financial year.

Thanks, Oskar. On Slide 10, our nonfinancial metrics. The total number of customers is up 13% to 1,547. Total interconnections rose 13% to 14,718, and total cross connects were at 9.5 per customer. On Slide 11, there's further insight into the diversity of our business. The breakdown of customers by industry shows strong representation from enterprise and connectivity as well as our system integrators and cloud partners. The increasing skew towards higher density deployments reflects the growth of hyper-converged infrastructure and hybrid clouds. On Slide 12, at 30 of June, 79% of installed capacity was under contract and 87% of contracted capacity was billing and generating revenue. In the past 12 months, we had a 24% increase in billing capacity, which now stands at 65.4 megawatts. With over 10 megawatts of contracted capacity at 30 June still to commence billing, there's scope for continued improvement in operating leverage and for strong revenue and earnings growth. Slide 13 sets out capacity and utilization. We now have a total plant capacity of 400 megawatts across the facilities that are either open or under construction. In New South Wales and ACT, we completed the fit out of S2, taking total capacity installed to its designed 30 megawatts. The S3 development is tracking well, and we're fitting out the first 12 megawatts, which is on target for practical completion in 2H '22. In Victoria, M2's expansion is also going well with 9 megawatts delivered in '21. And billing expansion works to further deliver 9 megawatts to support customer expansions. At M3, we've increased the planned capacity to 150 megawatts. Groundworks are well progressed and the building construction has commenced. We're on target for practical completion in 1H '23 with an initial 13.5 megawatts of capacity. And lastly, we secured a fantastic site in Western Sydney to support our long-term development pipeline with expansion capacity of at least 300 megawatts. On Slide 15, we provide a summary of our ESG highlights. As data centers continue to grow as a major feature in the infrastructure landscape, the focus on energy usage and the environment is critically important. During FY '21, the company made significant progress across numerous ESG areas, including energy efficiency, renewable energy generation, carbon neutrality, minimization of general waste, a reduction in water waste, giving back to our communities, driving social engagement and managing our supply chain. On Slide 16 is a summary of our safety highlights. As our national fleet of mission-critical infrastructure assets continues to grow in size and complexity, so too does the importance of keeping our workforce safe. With more than $0.5 billion of capital works and construction projects concurrently implied around the country, there is significant focus on achieving our goal of 0 injuries in the workplace. This includes programs to support operational safety, construction safety and most recently important priorities for COVID management and mental health. Turning to Slide 18, which provides our guidance for FY '22. We expect revenue of $285 million to $295 million, representing an increase of 16% to 20% on the recent record results compared to '21. Our revenue guidance is underpinned by strong growth in recurring revenues and long-term customer contracts with substantial contract capacity yet to commence billing. We expect underlying EBITDA of $160 million to $165 million with scale and earnings growth continuing to be driven by our generation 2 facilities. Total CapEx for the year is expected to be between $480 million and $540 million, with significant investments into our M2, M3 and S3 projects. FY '22 will be a year of continued growth for the digital economy and supporting digital infrastructure growth for NEXTDC. The foundations we've put in place over the last decade will see the company continue to scale rapidly through the next decade. We're building tomorrow's digital infrastructure today. And with that, we'll open the line to questions. Thanks, Travis.

[Operator Instructions] The first question comes from Garry Sherriff from RBC.

Craig and Oskar, the first one, just on that CapEx. It looks like it's clipped about $100 million from FY 2021 into '22. Just wanted to try and get a sense, is that COVID impacts from construction? I mean, does that push out any of your plans for developments for any of your assets?

Thanks, Garry. The $100 million in '21 CapEx was largely timing. So really, as the projects get larger, obviously, if you impact a few years, a big project and a large CapEx year might have been sort of $300 million as we're at the $400 million to $500 million type range, these large-scale development programs really just comes down to how much CapEx is sitting with at any given time. So the actual total amount that we had forecast in '21 will continue to flow through. It will just flow through into '22. As it relates specifically to COVID impact, obviously, right now, both in New South Wales and in Victoria, there is a reduction in the number of workers that can be on the site at any given time. Like most other people, it's a concern to us in relation to the ability to freely move around and get access to trade and other things that are needed to deliver large-scale construction projects. However, the forecast that we have given in relation to the delivery of S3 will only be impacted by a relatively short time frame. So as it currently stands, based on the government's advice for vaccination levels being achieved and people going back to work and us delivering S3, we would probably estimate something in the order of 90 days impact. Now that's based on current information. It won't make any real material difference to the current forecast that we have for the opening time, but we continue to watch that situation very closely in New South Wales and in Victoria as the numbers continue to climb. So not a significant impact at this point in time. The sites are still active. It's mission-critical infrastructure. We are still able to develop even when much of the rest of the industry in construction won't be able to continue. But it is something that does concern us, and we will continue to pay a very close attention to that, Garry.

A couple of more questions. One on the revenue per square meter and revenue per megawatts, which were down on the first half. Is that mix shift from hyperscale and how should we think about -- or should we still think about dilution on that -- on those metrics in relation to the hyperscale weighting across the portfolio given that's going to increase over time?

Yes, sure. Oskar, you can take this one. You might be on mute, Oskar.

Sorry, just taking myself off mute. Hi, Garry, thanks for the question. Look, there's really a couple of key drivers of that. The first one is really power prices. We do have considerable power pass-through revenue, particularly in relation to our larger customers. We have been able to negotiate considerably lower power prices, particularly in this calendar year relative to last calendar year. So overall, there has been an impact from that. Albeit from a net revenue point of view, we're still tracking very well. The other impact is really sales mix. We have had considerable large contracts go online over the past 12 months or so, and that has caused an impact in terms of the annualized revenue per megawatt. But as you know, those larger hyperscale deployments are much higher density. So we are still seeing an overall upward trend over time on a per square meter basis.

Okay. Understood. That makes sense. And the last question, just in relation to current projections, do you think you guys will go dark between now and when S3 comes online? And by that, I mean, I recall you saying there wasn't a huge amount of capacity to sell in Sydney between the S2 completion and S3 coming online. So I just wanted to get some form of color on that, whether -- where perhaps enterprise customers potentially going to competitors, if that is the case?

Thanks, Garry. In relation to S2 inventory, obviously, we've sold quite a significant amount with most of that being the entire facility and the target date to open S3, we were going to be very close in the context of inventory management, sort of seamlessly transition out S2 into S3. As I mentioned earlier, the construction progress and issues in relation to COVID management in New South Wales, we continue to watch very closely. It is likely that we will sell out. We have a very, very strong pipeline, obviously, of deals that we're closing out in S2, and that will well truly spill and spill S2 into S3 and COVID is an interesting circumstance. So whilst we continue to operate, getting people, particularly for international customers, having them move out and send infrastructure into the country and then have it installed and operational, that's creating interesting dynamics. But as far as our overall delivery of inventory is concerned, yes, we hope that we will be able to move customers nicely into S3. But we've got a very, very big order book that we're working through now. And most customers are very happy to engage on the idea of potentially they may have looked at deploying S2, potentially they can go into S3 early. The proximity of the facilities is in the same availability zone. That's a fantastic transition for us to be able to move clients seamlessly from S2 to S3. But at this point in time, our anticipation is that we should hopefully be able to transition from S2 to S3 without any customer impact. But I will note that we do have our strongest backlog of opportunities that we've ever had. And it is reasonable, and I would assume that we could sell out early, for sure.

The next question comes from Kane Hannan, Goldman Sachs.

Just a couple from me, please. Maybe just in terms of the revenue outcome sort of the bottom end of that guidance range from February and then as you're talking to that step down revenue per megawatt, just give us a little bit more color around what happens in February when you did upgrade that revenue range? And I suppose what's been playing out there, please?

Yes Oskar, you can take that.

Yes. No problem. So firstly, thanks for the question. The first thing I would note is we did come in into the revised guidance range. So no real surprises there. Every year, as we look at revenue, there are a few different moving parts. We do participate in some energy programs that can have an impact on the revenue from one period to another. So there was a little bit of a swing factor on that side of things. But overall, we came in within the range. So it's within the expectations that we had back in February.

Yes. I'll just add to that. Yes, the other thing I'd add just worth noting and something you may or may not be across power pricing to the extent that we monitor it, but power costs fell. So power costs down and net margin increased. And yes, it is what it is. Gross margin has improved. So overall, the movements in power pricing have been positive for us.

So no delays or any in terms of ramping up some of the installed base that we should be thinking about? It's just a greater-than-expected fall wind in power prices?

Yes. So as Oskar mentioned, obviously, when we upgraded guidance, both upgraded revenue and upgraded EBITDA at the time, we didn't necessarily forecast that power prices were going to fall quite as much as what they did. It's a bit of an art and a little bit of a science at the same time forecasting power prices, as many of you probably know. So yes, we had strong growth, and we upgraded into that growth. But we had the benefit of power prices coming down. And we don't sort of look at the variability on the revenue side as a negative in that regard. If the margin is improving, quality of the business is improving, so therefore, that's a positive. The revenue line can be a touch outside of our control as it relates to the final position of power prices at any given point in time.

Yes, that makes sense. And then just one last one. Just the S4 hyperscale campus. Just interested if you could talk a bit more high level around that announcement, so how we think about the funding of what's going to be a very large build? And then given that move into the pure hyperscale market, how you leverage your enterprise ecosystem to potentially improve some of the returns of that facility? Or just how we think about the returns of S4?

Yes. Look, this is a huge opportunity for us. We've not been, I guess, a key player, a key competitor focused on exclusively hyperscale opportunities, many in the industry that are just focused on winning in that very large scale space. As a company, over the course of the last 10 years, we've always been focused on those ecosystem-enhancing deals, the enterprise and government sector interconnection is a core pillar to our strategy. They're all margin enhancing. They drive a lot of outlier value. When we built very dense enterprise and government ecosystems and building, it's one of the key drivers why we have the largest number of cloud on-ramps in Australia today. If you're a public or private cloud hosting provider, you want to go where your target audience is. And so if you're a big public or private provider and you want to offer low latency sub-millisecond services, data doesn't need to leave the building. And the more you deploy your cloud on racks into those locations, the better number of target customers you have to access and the simplicity of those people accessing that infrastructure behind the network, so to speak, offers a lot of advantages. So that strategy has obviously served the company very, very well over time. We're not changing that strategy. We still see interconnection as a core pillar of our growth over the next decade. It's a very, very important piece of the puzzle. Connectivity is the glue that binds the Internet to public and private cloud. Hybrid will continue to be a feature for a very long time. So as it relates to S4, we have an opportunity now to think about our growth pipeline and where the hyperscale market is going over the next decade. If you consider in each of the availability zones, larger providers will have somewhere in the order of the 300- to 500-megawatt target capacity each, that would put the total zone power target in IT load somewhere in the order of 1 to 1.5 gigawatts. So if you are looking at a handful of facilities in Western Sydney serving the hyperscale market that are in that 1 gig to 1.5 gigawatt target range, we'll have a good sized facility that can take a similar amount of market share to what we would have in the enterprise business today in Australia. Obviously, I'd like to see us improve our market share, but the key for me is using our key points of differentiation and our competitive advantage as a company. We built Tier 4. We certified with the Uptime Institute to operational Gold standard, the highest operational standard in the world today. We're the only operator that's NABERS' 5-star certified, which allows us to independently certify and guarantee to our customers they're getting the lowest power costs, very important going forward, and obviously, the ESG implications that flow with those for sustainability, water recycling. And then you've got all of the ecosystem elements. So we will take our network, as we have done with S1 and S2 connecting to S3, we'll then have diverse metro fiber connected to S4 that will allow us to offer our enterprise and government customers campus solutions. So the enterprise and government component of the business just because we are looking at this facility primarily as a hyperscale campus, it's going to be extremely attractive to enterprise and government customers who can tap in to those network services. So today, more than 730 partners and tens of thousands of behind the Internet connectivity services puts us in an extremely powerful position to leverage the size and scale of our network. So the S4 opportunity is just an amazing opportunity for the company over the course of the next decade. The sizing of it is in a similar fashion, when we look forward to total capacity over the next decade, a similar market share to what we would have today. Obviously, I hope to continue to grow that and the mix of customers -- the things that we do that are unique in the industry will continue to focus on as our competitive advantages. So margin is obviously published globally. People know what hyperscale returns look like, and that particular point gives us the opportunity now that we have secured the land to work with potential funding partners to ascertain if that's the right long-term balance sheet structure for the company. If you look at successful examples, most recently, Equinix's partnership with GIC, for those who may have watched the Equinix Investor Day, you would have seen Keith talk about how outstanding the performance of the xScale business has been. GIC is a partner in their funding capability. I think all of those things are very logical, and we would expect to explore a similar path to that. We've got good time to plan for that. And large scale deployments on this -- of this size and scale take a couple of years. So we're already actively working with our customers on what that planning would look like. And I'm just thrilled that we're in a position to be able to go into an entire new market that expands the company's execution capability adding to a new zone,and take all of our competitive advantage, things that the company has been recognized for doing well on a global scale and do that over the course of the next decade. So S4 is really exciting.

The next question comes from Paul Mason from E&P.

Just a few from me. So the first one, just on -- towards the back of your presentation, there's sort of a bullet about S4 saying 6 further developments to be disclosed, is that something like just S4 is going to be built as like 6 by 50-megawatt parcels or is that something else?

Thanks, Paul. I can always count on you to find the detail in the presentation. So thanks for picking that one up. Yes. So obviously, there is a long pipeline of facility development work that we've been doing over the last couple of years. It took a couple of years to get S4 out given the size and scale of that. It's a huge, huge development. And when you're developing over 100,000 square meters like that, it is years' worth of planning. There are a number of other developments, obviously, concurrently going on in the background and land that we're acquiring to continue to expand our footprint. S4 was the first one -- first new additional one that we have disclosed. But there are a further 6 that are currently in the planning phase at the moment that will be announced shortly. So there will be a number of additional new announcements for new developments over the course of '22, that will see us bring another [ 6 out of 10 ] developments to market.

Right. Okay. Great. Second one is just on M2. I just wanted to make sure my numbers are right. So the 9 megawatts of extra capacity that you're bringing forward, that brings you to 28 that you'll have fitted out at that facility? Is that right?

Yes. Oskar, do you want to take it?

Sorry, Paul. Would you mind repeating the question?

Yes. So in the presentation, it says that you're bringing forward another 9 megawatts. And the way I read it, so there was like 10 megawatts historically, you've delivered 9 this year, and it looks like you're doing another 9 next year. Is that correct?

At M2, that's correct. Yes.

Yes, 28.

Yes, correct. Okay. Cool. And just you sort of skipped over the bullet on P1 going to 10 megawatts. I was wondering if you could maybe talk a little bit about what's going on there?

Another good pickup, Paul. Yes, there's another -- the final design expansion of P1 will be 4 to 6. So we'll take the conservative position at the moment. We have acquired the additional surrounding land at P1. P1 is down to its last [ data pool ]. We've got -- the team has done an outstanding job in Perth, obviously, with P2 coming on, having 2 active locations to sell, which is our strategy. We want to have 2 active inventories to sell in every market so that you can win both ends of deployment, both for enterprise and government and for hyperscale. Being in a position to have inventory in both markets means that we need to obviously continue to edge out in the existing facilities and certainly for the first generation ones. Many of those, obviously, as you know, are full. So we successfully acquired surrounding land for P1 to grow. The current design will see us double the size of P1 potentially. So it was 6 -- adding another 6, but we're taking the conservative position of calling 4 at this time until development approval is worked through with counsel and that would take 1 to 10, potentially 12 megawatts, and that will give us a like-for-like inventory to continue to sell P1 and P2 in tandem. So a good pickup, Paul.

And so just the last one for me. Just a more strategy question. So historically, you guys have had a policy that you don't build speculative wholesale capacity in advance of an order, the build of the fit-out follows the order. Is that -- that's still on foot, that policy?

That is true.

The next question comes from Tim Plumbe from UBS.

A couple of my questions have been asked already, but just one that I was hoping you could elaborate on, please, Craig. The 5.5 megawatts that you added this year, of which 4.5 of those were in the second half. Can you just talk to the composition of those? And Digital Realty recently noted that they've seen an uplift in terms of enterprise customers coming through and kind of putting that down to that post-COVID movement. Have you guys started seeing that accelerated trend coming through in terms of enterprise customers? And do you think that, that plays out to a further extent in FY '22?

Thanks, Tim. So yes, look, it was obviously a good 0.5 megawatt. We're on track to look for that megawatt worth of enterprise and government in the chart. Remember, that's a term that's the final number, puts and takes. Customers go up and down, consolidate, move to the cloud, they repatriate what goes on inside the deployment is incredibly dynamic. I mean, the volume of movement around in platform architecture at the moment is just like nothing else I've seen in my entire career in the IT industry. I've been doing IT for 25 years. And I've just never seen anything quite like what we're seeing at the moment. It's just amazing. And because of that fluidity in enterprise architecture and people sort of experimenting public and private cloud and then repatriating some data sets and got obviously very large-scale legacy infrastructure that's causing hybrid deployments, the enterprise is probably the single greatest opportunity that we've got today. Hyperscale is great. It's big. It's exciting. Everybody is drawn to the big infrastructure like opportunity. That's why every global sovereign wealth fund in the universe is chasing us and others to get a seat at the table. Everyone wants to play in the space. But the real opportunity, the great quality business is continuing to work with enterprises that are managing their own infrastructure and utilizing public and private cloud, where it makes sense. So we did see very, very strong enterprise growth in the second half. That will translate over time. It will continue to translate over time into growth for S2, S3, et cetera. But Tim, the key point for me really is that the performance for us in the second half was characterized by a couple of really interesting trends that are emerging. The first one is that we're seeing what we would call the next wave of challenges in the hyperscale space. Now they may not be hyperscale in size or what you would expect a hyperscale to represent. So remembering that Microsoft, Amazon and Google account for 70% in the market. So today, if you're at 70% of the global hyperscale business, when they're placing an order, that order could be 3, 5, 7, 10 megawatts, and they're very good sized orders, particularly given the volume of those that they're placing in multiple locations. So as an example, one 10-megawatt order in a single location, we'll have 2 or 3 other 10-megawatt orders that have dragged along behind it. So a 10-megawatt order in one of our facilities actually will result in 30 or 40 megawatts worth of deployed capacity. But what we've seen, which is the really exciting part is the next wave. So we did get some wins in the hyperscale space during the second half that were challenger brands like OVHcloud, Europe's leading private cloud provider. So the OVH team, obviously, that's a big challenger opportunity to the Microsofts and others of the world. And there are many more of those coming. So the pipeline is really booming in that regard because the challenger brands are starting to edge out into regions. And on a global scale region, Australia is a region when you think about, if you're in the U.S. or in Europe, you're edging out into Australia. When we talk about regions, we're going from Sydney and Melbourne talking about Brisbane and Perth. There will be more markets. We will edge out further, and we'll talk about edge strategy soon. Not today, but we will obviously reveal a little more about our edge deployment strategy and where regional and edge data centers will go because it's a significantly important piece of the digital infrastructure puzzle over the next decade. So playing in the edge will be important pillar about data services strategy. But the key point to call out, Tim, to your question is that challenger clouds are starting to emerge. And in the enterprise, we're getting continued growth as a result of COVID. And Tim, I can use your own company as an example, and thank you for your business that we managed to secure UBS globally as a leading financial services brand. So we'll be hosting your organization as a client in Australia. And there are many other global banks that still have their data centers on-premise in their offices, and COVID itself has been an important catalyst to driving transition and the prioritization. So we found a compelling event that in a lot of cases enterprises were missing, they didn't have a competitor compelling event, the catalyst driving them to prioritize getting out of the on-premise data center. And COVID has been a really good reminder for organizations, particularly on the risk management side that they need to be able to operate remotely. So that's a long answer, Tim, but there's quite a little bit in there to unpack because there is a lot of things that are starting to emerge in growth that we really haven't seen before. So it's super exciting for us.

Completely agree. And so I mean, you guys have got pretty good line of sight in terms of deployment of hyperscaler capacity into your data centers. I guess one of the areas that can surprise to the upside is the uptake of enterprise. Have you factored in much of an acceleration into that FY '22 guidance? Or is that an area that could provide upside if you do get that accelerating trend coming through in '22?

Yes. Look, we always expect -- I get asked the question regularly, Tim. Why don't you expect the enterprise number to be bigger or faster? And it's not that we don't want to win a large share. But we continue to focus on the premium end of the market. We're not the price leader. We're not trying to sell cheap. We are the premium product. We are Tier 4. And at that level, we are focusing on segments of the market that are premium end. The market is larger than everything that we claim. If we wanted to go downstream and double in other lower-cost data center services, we essentially could in the future, but it's not currently a feature of our strategy. But the enterprise, when people look at the size of hyperscale orders, they sort of see big numbers and go all that sort of that's where all the focus is. But when you think that we win racks literally every single day, 2 racks, 5 racks, 10 racks, day in and day out, that enterprise business is the high-margin business that drives outline margin performance for the company. It drives high-volume inter-connectivity, and it is incredibly difficult to unpick over time. So the real focus for us to continue to build great profitability in the company long term is that highly diversified mix of enterprise customers and the very rich interconnection ecosystem, and that's a wonderful margin that comes along with that. So in the enterprise, yes, we want to continue to win our own fair share. Hundreds and hundreds of deals are done, almost a couple of hundred net new customers in FY '21. So yes, sometimes the deals are a little larger. But every day, a rack a day keeps the CEO away, and we keep driving this up team to win our own fair share. But yes, enterprise is exciting, and it could certainly surprise on the upside, Tim, but we always take a conservative position that a megawatt of run rate is a very strong result, and we'll continue to do that. This also need a change. The only thing that we lose in our power upgrades and downgrades and other bits and pieces, change is a relatively small feature in our business. But a decade on, we're still focused on the enterprise.

Got it. And just last question from me. Microsoft have got a couple of DA approvals out in the Sydney market at the moment, which is a bit of a change in terms of the way that they've done things compared to historically. Any comments that you can make in terms of how that could change the dynamics of the industry?

Yes. Look, I think that past history is the best predictor of future performance, but just [ 10 of the 16 ]. When you look at what Microsoft and Amazon and Google generally do in the U.S. market, it's close to home, as the size and scale of the platforms get larger, you would expect them to do a little bit themselves as they have done, Amazon have been building their own and using Palo in Australia since they started. Microsoft should do the same. Google haven't, but they could do the same. I mean these are, I guess, just features of an installed base that's 10 years mature. There's a baseline level of capacity that if you're building and thinking long term, and these organizations, just like us, have to think long term, you've got to take a -- at least at a minimum, you're taking a decade long view after the deployment of your infrastructure. The bulk of these contracts that we sign are decade-long with decade-long options. And in the majority of cases, they're not moving in situ infrastructure because once it's deployed, it's carrying the base load. So speaking everything that's being built, it's being built to try and offset some of the additional demand variability. Most organizations in that regard still struggle to forecast their own future demand. As the success, you would have seen the extraordinary numbers coming out of all of those key CSP leaders. And it's just a great feature of their business. I mean the business gets larger, the volume is online products, online gaming is becoming a really big feature. We've won some deals. Just as an example, Pentanet in Perth as a partner, we're deploying their gaming platform, which is the video gaming platform that they exclusively distribute between Perth and Sydney. These are all features of, I guess, just maturing growth, both in private cloud and hyperscale, and you would expect these guys to do a little bit of that as they've done successfully in the U.S. So I don't see really any change in strategy. I think that's consistent with what they've done in other markets. And in the most part, we're all focused on trying to keep up developing and meeting that exponential demand curve for the next decade. We still don't build ahead of capacity. So one of the challenges for everyone really is, you don't know where the designs are going. If I just take M2 as an example, the M2 design, the level of density, our changes on the UPS infrastructure and how that works changes on airflow and air management, a lot of the PhD research work we're doing in relation to cooling technology, it's all -- it's just amazing stuff. But it's moving so quickly that you don't want to build a lot of capacity in advance as having sold it. We want to build enterprise and government capacity because you need to build it on hand, you need a little bit of inventory on hand to meet customer requirements, so you don't go dark. But in the hyperscale space, whether it's public or private cloud and regardless of the size and scale of hyperscale, a lot of those wins that we had in the second half are, what I would call, the mini hyperscalers or the emerging hyperscalers, the key challenges in that market. And just because they're small doesn't mean that they're not incredibly innovative and the type of designs and technology that they're building themselves. They're building their own GPU stacks and storage -- flash storage arrays and other things. And each time we do a unique design, there's just so many elements that need to be customized. It's a key feature. It's a competitive advantage for us having our own in-house engineering team to do these designs in collaboration with customers. It also earns a lot of credibility for us as a partner because we're not going after third-party engineering organizations, our Chief Engineer and the engineering team that we've had, obviously, as a key competitive feature of our business from Day 1 continue to develop very deep relationships inside the research and development engineering teams with our key customers. So yes, I think it's an interesting dynamic. It's a good feature of the industry. Not unusual or unexpected, and we're all focused on where we need to be to support growth over the next decade.

The next question comes from Bob Chen from JPMorgan.

Just a couple of questions for me. Recently, you guys put out the announcement that you've got the government or public certification. Can you talk a little bit about that part of the market, the sort of the public sector part of the market? And what sort of capacity could you pick up there?

Thanks, Paul. Look, this thing I think is really exciting. So for anyone that's not across the detail there, federal government, obviously, very focused on changes in the threat landscape. So foreign actors, the underlying ownership of Australian infrastructure assets. And clearly, in the past, if it was ports and power infrastructure and other critical services, that would have played a role and being involved in the ownership considerations related to those. But in the data center sense, what has collectively become known as the Global Switch cause. You're in a situation, a circumstance where what was a U.K.-owned business becomes owned by the Chinese. And when it's 100% owned by the Chinese, the concern is obvious in that regard, the FireEye partners and the FireEye defense partnership alliance that Australia is a key player in. That is an important feature of protecting Australian critical infrastructure. So the DTA, the Digital Transformation Agency, which is the primary agency for the federal government in this area, as a result of a lot of the questions that were being asked on ownership, Bob, if you haven't seen it, Bob Katter was regularly asking Scott Morrison in Parliament, why the Department of Defence and tax office and others weren't urgently required to leave Global Switch as a result of that Chinese ownership. And Bob Katter kept pressing on the Prime Minister as to how important it was to get any federal or state-owned IT infrastructure out of Chinese-owned facilities. So it's been a very, very hot topic. For us, the DTA certification at certified strategic, there are 2 levels. Certified strategic is the highest level. And having gone through that work with the federal government for many months, we took the decision -- rather than to certify maybe only 1 or 2 facilities, we took the decision to certify the entire national fleet. So it took some time to, particularly with COVID, the federal government auditors needed to visit every facility in the country for us to get that certification. But we now have the single largest fleet of federal government certified facilities in the country today. They are certified to the highest standards, certified secure. There is no higher standard today available. So I'm really pleased that we're in that position to be a key provider to federal, state and local government agencies and alleviate any of those concerns. The issues in relation to foreign active security concerns are real. I spent the best part of the decade in security industry before I spent the last decade here at NEXTDC and nothing has changed. The threat landscape just continues to evolve and develop and become more sophisticated. So I think the government's decision to play a heavy handed role in ensuring they can protect Australian critical infrastructure, particularly in the digital age is a really important one. So we're thrilled, privileged to be certified strategic, and I see that as a great opportunity. We want to win more in federal government, the federal government markets, a couple of hundred million dollars a year, and we'd love to be winning our own fair share there. And we're going to continue to do more in that space and try and grow our current footprint there. So it's a great opportunity. Pleased to have certification out. And yes, it's a big segment of the market that we'd like to win more in.

Okay. Great. And then obviously, a lot of the construction going on and even further expansion capacity going forward as well. I mean how are you thinking about sort of the funding requirements for some of this expansionary capacity in the future?

Look, we've got a lot that fire power at the moment to be deployed in the existing developments. The team will be upsizing our debt facility, not that we -- we've certainly got plenty of cash on hand at the moment, and got a significant amount of debt available to us to continue to invest in the business. But we've already kicked off the process to upsize our existing $1.85 billion debt, very, very strong reverse inquiry. The level of inbound encouraging the company to substantially upsize the debt facility. The movement in financial markets, credit pricing, all very positive, and the company has been presented with the opportunity to substantially reduce its cost of debt and significantly increase the size of debt. So some number in excess of a couple of billion, and the team are working through that at the moment. So I would expect that over the course of the next 2, maybe 3 months, the team will work with those that have been inbound to encourage us to take a substantially larger secured debt position in the market to take advantage of significant pricing improvements to reduce the cost of our debt. And it's just a great time. Obviously, this is perfect timing for us to get a really good reduction in the cost of the debt and improve our overall WAC. So nothing like getting free money. And when there's free money on offer, I'll be the first in the queue with my hand out.

Thank you. That does conclude the time permitted for questions today as well as today's call. Thank you for participating. You may now disconnect.