Welcome everyone. I would like to begin today by acknowledging the Gadigal people of the Eora Nation, the traditional custodians of Tulawalada, where we meet today, and pay my respects to elders past, present, and emerging. I also extend my respects to all Aboriginal and Torres Strait Islanders with us today. Welcome. I am Alison Gerry. I am the Chair of Infratil, and it is lovely to see so many of you in the room today.
It is a great room. We may be missing the views of the Harbour Bridge, but I am sure we will adjust to that. It was also great to see many of you at our welcome drinks last night. For those of you who may not yet be investors in Infratil, I thought I would start with this map, which shows the global diversity of our asset portfolio.
Some of those dots may look like they are in the Pacific Ocean, but they are actually in New Zealand. We do receive feedback that Infratil can be complex to understand given the number of assets, and we are working on addressing that by refining our portfolio. We also acknowledge that Kiwi investors have had an advantage. They have had 32 years to get to know us. In the spirit of Trans Tasman Corporation, we have decided to hold our investor day for the third time in a row in Sydney.
One of the great things about Infratil is you have a great bunch of Kiwis in the room today, including many of our New Zealand investors, who are happy to promote our Australian businesses and also impart their knowledge. Infratil would really welcome more of you to join our success story and be supporters of Infratil as shareholders.
You might be able to see from the slide, I cannot quite see the angle of that, but hopefully it shows you that more than half of our portfolio is now based in Australia. CDC is our standout business, and they have had a stellar year, and we are going to hear more from the team later today. Moving to our board. We have six of the Infratil directors with us today in the room, so please do feel free to introduce yourselves to us or ask us questions in the break, and I might just get the Infratil directors who are here to stand up so you can spot them.
Great. We have got Anne, Andrew, Kirsty, Brad, Jason is here somewhere, but Paul Gough, our London-based director, cannot make it today. Also our newly appointed director, Tiffany Fuller, also has a previous commitment.
Peter Springford retired three weeks ago at our annual shareholder meeting after almost a decade on the Infratil board. Both Tiffany and Brad are Australian-based directors, and we are really looking forward to having their perspectives around the board table. This slide is a summary of what the board calls our strategic KPI dashboard, and it illustrates the key areas that the board is actively monitoring.
First and foremost, and you have heard me say this many times before, our cornerstone goal is to deliver shareholder returns between 11% and 15% per annum after fees and taxes over a 10-year period. This is based on share price growth and assumes that dividends are reinvested. We are really proud to have delivered more than 18% per annum returns over the 32 years since Infratil's inception. Our performance compares favorably to listed infrastructure companies that we are benchmarked against across multiple time frames.
I might draw your attention to a recent executive summary of a report that the Infratil board has commissioned on a benchmarking exercise that you might like to have a look at. It makes for great reading. Another measure that we follow closely is our calculated net asset value or NAV per share after fees. We know there is typically a discount to NAV because market views on value can differ from the independent or market-based values that we use for our portfolio companies, and we are seeing a little bit of that this week for sure.
We also know that we can do more to help markets understand the value in the longer-dated Infratil valuations. This is a large reason for us being here today, so you can hear directly from us and hear from our portfolio companies as to what we see ahead of us.
We are also continuing to enhance our disclosures. CDC's shift to reporting on leasable and future contracted capacity is a good example of that. We are always happy to hear from you if there are any other areas for improvement. We know investors are cognizant of asset and sector concentration risk, particularly with CDC's strong growth, and the board monitors this risk. We have spent quite a bit of time stress-testing CDC's valuation. Last December, Infratil received its inaugural BBB+ credit rating from S&P Global Ratings, which we are really pleased with and a little bit surprised with.
The rating recognized the strength, quality, and resilience in our businesses and has provided us with very clear credit metrics for us to operate within. ESG reporting is another focus area because decisions grounded in responsible stewardship are part of creating long-term value and managing risk.
We are very pleased that we have been recognized as leaders this year in various external assessments, particularly I think it was Morningstar Sustainalytics. Finally, on this slide, portfolio companies, the relationships, and resourcing are an area we are paying closer attention to. Greater collaboration between our portfolio companies is something that could unlock more synergies and value, and you are going to hear more about that today.
Turning to the agenda, we want this day to be interactive and help everyone's understanding, so there is going to be lots of opportunities for presenters to ask questions throughout the day. Jason is going to kick things off with an update on Infratil's portfolio strategy. We are then going to hear from the Morrison team of Paul Newfield and Will Smales on how Morrison's own growth supports Infratil through asset management and new opportunities.
After morning tea, the Longroad Energy team will give us an update on how they are moving to capture value from the exceptional demand for energy in the U.S. We will break for lunch, followed by One NZ's update on the New Zealand telco market. We have Nick Judd, our incoming, newly appointed CEO, who is going to give us his perspective. The CDC team will then take us into the fast-paced world of data centers. We all know we could spend many hours on this topic, but we are aiming to wrap up at the latest at 3:30 P.M. With that, I would like to hand over to Jason. Kia ora.
Righto. Good morning, everybody. Nice to see you all. Thank you for coming along. I have been looking forward to this. What a time to be alive, right? I feel like when I get down from the stage, it probably will have changed again. But if you, like me, thought you were in an alternate reality or woke up and were dreaming, seeing the president calling Jensen Huang on stage at the All-In Summit yesterday morning, you will be as excited as I am to find out what happens tonight. I think I should thank you for being here rather than the All-In Summit.
Maybe we should move it and go there next year because it seems to be the center of our world at the moment. But we are much more modest than that. We do have things to say about AI slowdown.
Actually, the thing I was most prepared to talk about, which we should still talk about, I think it is probably more meaningful, is rising long-term real rates, which are particularly relevant for long-term contracted infrastructure, like the stuff we build. But let us get going. I wanted to start with a couple of introductions, or three actually, and then I will get onto the portfolio strategy update and those fun topics I mentioned. We made some recent changes in the kind of executive structure, I guess, at Infratil, at the start of the year.
The idea was really to create dedicated roles focused on execution across what is now quite a big distributor portfolio, finance, which has become bigger as well, and also investment. So we have Andy Carroll, who was our CFO. I am getting an echo here. Anyone else is?
Who was CFO, who is moving into the COO role. His job really is to set and reset the priorities across the group and then ensure that they are executed in kind of the right sequence, which is a material job now. Then we have Matt Ross, who was Deputy CFO, and we are delighted he could step up into the CFO role, which means we do not miss a beat in that seat because he has a long history with Infratil, longer than both Andy and I actually.
Lastly, we created this new role, which Lee Coker took, focused on investment, and his job is to curate the pipeline of both internal and external investment and reinvestment opportunities so that as a board, as a leadership team, we can rank those next to each other, put those next to our funding availability, and come up with clear direction back to our teams and our portfolio companies about how we want to allocate our ultimately finite capital. So those people are around the room.
They are versatile players and coaches. They draw extensively on the larger and deep Morrison team to actually get stuff done. But these guys, alongside me and a few other people, are the kind of fulcrum points of all that. So if you see them around, they are important people. If you have got investment ideas, just bug Lee.
If you have other questions, bug the other guys. Now strategy. Last year, we put up this slide, which showed our portfolio strategy divided over three pillars. You had cash flow and then mature and future growth.
The three, I think, key takeaways that we wanted you to get last year were this idea that we saw going forward, focusing on a smaller number of scaled, maturing growth engines, CDC and Longroad, data centers and renewables in particular, which would be supported by divestments of less core assets, ones that we could not see scaling in our portfolio, both to create that refinement, but also the funding required to fund their growth, which was a big question on people's minds. That was point 1. Point 2 really was that we saw the pathway to getting our operating cash flow back in balance to make the dividend sustainable.
It is an important piece, I think, of the Infratil story. We are not necessarily trying to grow it, but I think making it sustainable over the long term has been an ambition, and I think we have made good progress on that over the year as well. Focus, divestments, balance cash flow, kind of three. What did we see during the year? We obviously saw progress on divestments of NZD 600 million done.
The last of that first batch still ongoing, and then announcing that further divestments were on the horizon as well. Another NZD 1 billion we talked about at the full year. That continues. We saw good progress on the operating cash flow metric, with excellent progress from One NZ, and they will talk more about their pathway to making that sort of level of distribution sustainable.
Then obviously we have great progress from CDC and Longroad, and CDC in particular on their operating cash flow metric makes it such a shorter route to free cash flow being available from them for reinvestment or distribution. But really underpins the main theme that we are seeing continue from last year into the years ahead that is driving our portfolio-level thinking, which is this. I think everybody has seen, and maybe we can flick through this, but we are seeing surging token demand all around the world, driven by the capabilities of the models getting better and better for coding, cybersecurity, all these sorts of things.
In parallel, token costs falling as open source models become available and people are able to shift their workloads as they get comfortable with where the frontier is at, onto more optimized systems to enable them to generate the tokens that are more cheaper, which again, drives more adoption, more token demand, and so it goes. There is no doubt model competition that is intensifying with that theme.
The frontier is challenged to keep pushing the frontier, to keep being the first point that people go to decide how they are going to implement AI. But there seems to be very real benefits that enterprises are experiencing through adopting it in the areas that I talked about and more, that will continue to drive the system over time. Perhaps now is a good time to just pause on this AI slowdown ideas.
Even with where the models are at today, their capability, that on its own will continue to require significant investment in infrastructure to serve that capability across all enterprises who can make use of it. We are not even using all of that yet today as well. I saw overnight, Satya Nadella, Microsoft's CEO, was saying more or less this at the All-In Summit as well.
Even if you pause the frontier, which is obviously pushing out in all directions, actually serving up and making use of exactly what we have now, will continue to require significant investment, particularly because of the next feature that I wanted to talk about more today anyway, which is the supply constraints that are in the system for building the infrastructure to serve these workloads. They exist all across the theme and all across the kind of supply chain.
Things like semiconductors, I think are well understood and obviously a big trade over the last few years. The thing that we see as staying constrained for a longer period of time is the delivery of power at scale to these data centers. That is because these power systems take a long time to build, and they always have. It takes a long time to deal with the grid to get your projects approved, there are all sorts of nuances to getting them financed and built, and sometimes they run late. Our central thesis is that at that layer and at the data center layer, that is a two-year, three-year lead time piece of infrastructure. Microsoft's CEO was talking about this last night. You do not stop that now, right?
Because the frontier might be flying to where it is going to push out by a few months, by a year. You actually have to build now in order to be able to deliver that capacity in two or three years time. This is why, right? You see massive demand infliction for data center capacity growth. This chart in the middle is pretty interesting.
I think it is a Goldman Sachs one, which shows even with what we are expecting to build today in the U.S. on the power side, they are still projecting an 80-gigawatt shortfall of power for the sector in 2030. To give you a sense of that, I think I was talking to Paul Gaynor last night, the best year that U.S. renewables industries ever had is 40 gigawatts. So that shortage is not going to go away. It is going to be short.
It is going to be short whether the models move the frontier or not, because we still have not got enough infrastructure to serve up what we have got to everybody who could benefit from it. It is our core kind of belief that is driving our portfolio thinking. In that sort of environment, this sort of second box on the right here, we believe that existing platforms with long, strong track records in developing responsibly in this space, which is an increasingly important requirement for having a long-term business with advantage access to supply chain and suppliers with deep pipelines of projects that can be connected in any time frame.
We have talked in the past in the U.S. about interconnection timelines being five, seven years, and we have developing projects now that we put in interconnection queues literally that long ago.
Unless you haven't been in this game for a long time, you're not going to get your projects developed, and it's getting that way in data centers too, I think. Those sorts of platforms in a scarce environment are going to be well-placed to obviously deliver good capacity, but I think also at least maintain strong returns on equity in this space as delivery costs continue to change, as this long-term interest rate theme plays out in whichever way it plays out.
This is where we see the returns generally sitting from our research at Morrison, and you've probably got your own. You see this trend of expanding returns on equity, EBITDA per NZD of CapEx, as you move right along the AI infrastructure value chain. Certainly, as you move to the right, well in excess of an investment-grade cost of capital.
What you also see is, as you're on the, in my view right, probably left on yours, at the neo cloud end of the circle, you get shorter tenor, basically. Shorter duration of contracting, which obviously goes with the shorter asset life, and then you're getting higher returns. In general, on the other end of the AI value chain, which is where we are at the moment, you're getting long contracted duration cash flows with good margins over your cost of capital that are very attractive, I think, to us as an investor who's targeting that sort of 15-plus zone.
It should be no surprise that we see a real advantage position for our existing businesses to continue to develop both in those verticals, if you like, in the AI value chain as we are today, but also to continue to build an adjacent options around each of their verticals as we see some of them converging. There's kind of three areas here I wanted to focus on. One is at that power and data center level, the two columns on the left. There's all those adjacent options that each of those businesses can, and I'm sure will be looking at, and you'll hear some of it today. If you're at the energy level, you can run more merchant capacity.
If you're worried that power prices and interest rates are going to go up and your long-term contracts are going to be stuck because you've got a fixed revenue line, then you can start running more merchant capacity, more firming and storage, and capture more of that kind of in-contracted period return increase to balance your portfolio. You can look at transmission extensions, which definitely the Longroad team will be looking at. On the connectivity side, we have EonFibre out in the fiber and connectivity space. On the data center side, we think about geographic diversification a lot, and I'll show you how we're thinking about that just in a minute. Then there are lots of different sovereign applications. My point before, we don't have the infrastructure to serve up everything we've got today.
A lot of this stuff still needs to be built in a lot of countries. Our existing position, for all the reasons I talked about before, gives us an advantage, a reason to win in those spaces, I think. The super interesting part, I think, just for now, or most immediate really, is this idea of integrating power and data centers together, which I think I talked about at the full year. You've seen a first example of that within the group with CDC and Contact announcing a joint venture to explore developing their site in Taranaki, which has been a fantastic way to think about, "Well, how can we reuse existing infrastructure? How can we site a data center in a place where it already had an industrial use?
How can we already have a strong story about where the additional power is going to come from to power this?" I would see more of those ideas as needing to be on our radar over time, and a great opportunity for the Longroad team as well, who've also looked at their own pipeline and will talk about how they see this developing in the U.S., which is a real theme. It doesn't end there, and this will change over time.
I think of that as kind of the most near-term focus at that end of the value chain. As this story rolls out, there will be more opportunities to the right, and some of them we're seeing already. Infratil has had a great history, I think, of buying incumbent businesses, transforming them, and then deciding whether you're a long-term holder or not.
We're seeing the ability to use AI and deploy it at scale within existing businesses that have strong defensive characteristics as ways of creating returns that you wouldn't necessarily ordinarily expect out of these businesses and One NZ is a very live example of that, and you've been hearing about their AI transformation story. They've revealed a couple of other ideas this year with EonFibre and RANCo as ways to kind of optimize your return out of that asset and optimize the way people think about them. Healthcare is very much live in this space as a major theme there, driving productivity. We would expect to see more opportunities like that in the future as well. That Infratil is still very interested in addressing. It's definitely part of our DNA and part of what differentiates us.
It'll just be finding the right time when the capital is available and when the opportunity presents itself to do it. I would definitely put that on the radar as a theme well within our capabilities long term. As you roll out into the future, physical AI is still a big focus area for us. We've been talking about automated logistics and things like that for a couple of years now, and this seems to be getting closer and closer. Will's going to talk about a couple of ideas Morrison's been introducing at a scale that's a bit small for Infratil now, but certainly we'll get the benefit of learning from those as more scaled opportunities in this space come out.
The big theme is not everything is AI, but we have an advantage position to build out into the AI infrastructure and beyond value chain, starting from our strong bases and momentum in power and data centers. You can see this expressed in the portfolio proportions. This is an updated version of it showing CDC in the portfolio, the energy businesses, and the others. You can see on the right, I talked about geographic diversity. Pretty good diversity, actually, in the energy space and as a thought on our mind in the data center space as well. Double-clicking on data centers. Greg will talk a lot about this with his team later on. We see strong growth, but also strong downside protection in the space, and we see them as being well-positioned to address adjacent options.
At the bottom there, you've got a graph that's showing the proportion of the independent valuation today that sits in existing operating data centers or contracted data centers that we're in the process of building, that CDC is in the process of building. You see over half of that valuation is really in existing operating businesses.
If you stopped doing any more business today, you're still on a path to over NZD 2 billion of EBITDA and an incredibly strong operating free cash flow if you were not investing at all, which I think is a great downside-protected position to be in. Greg will talk about how they continue to have a really highly differentiated platform on multiple metrics to be well-positioned to continue to develop at attractive returns on equity, supported by a strong funding platform. Kao Data's there. They've had their own strong growth.
Different business, but actually addressing an interesting, slightly different niche from CDC, providing some geographic diversity as well and some good prospects ahead. On the energy side, clearly the U.S. part of the business is accelerating the most strongly. Again, strong downside protection, I think, in this business as well, with more than 50%, well over actually, of that business represented in existing assets that are long-term contracted or assets that we're building now that are long-term contracted. In this, the contracts are comprehensive. You've got a long-term contract over 17 years on average. It's covering 96% of your revenue, and the counterparties are A-plus rated on average on the other side of these. That creates an incredible base from which the business can grow in all environments.
They're seeing strong growth and acceleration as well of their own and a little bit of an upgrade from the internet in terms of what their outlook is. On the other renewable energy businesses, we see those as developing longer-dated options in exactly the same place over time. They're adjacent to Longroad Energy as the kind of core, if you like, and creating longer-term options as different markets catch up with the U.S. on developing their AI infrastructure over time. There's a good spot to talk about the whole real rate thing. Probably in the portfolio, the Longroad Energy valuation is the most sensitive to long-term rising real rates, particularly if that is persistent, and there are reasons to believe that could be.
To give you a sense of that, this is because on the operating side of the business, you have these long-term contracts, and if rates go up, you cannot really do anything about creating more revenue very easily. There are some ways. To give you a sense of it, I think a 50 basis points increase in risk-free rate is about NZD 160 million impact on the opco valuation.
So it is about a 4% movement in the valuation. So it is pretty tough. To give you a sense, that is probably also about a third to a half of a year's work on the development side of the business. So you can work really hard and do what you said you were going to do, and then you get 50 basis points on the rate and your valuation is flat. So that is a little bit what it is.
There are two counterbalances to that that are natural within the business. One is, theoretically, if interest rates remain higher for longer, then the merchant power price should increase as well because the cost of all the kit that is being put on the system is going to be higher. Unfortunately, that takes a long time to come through merchant curves and a long time to turn up in your valuation. But that is, in theory, what should happen, which is why I said, well, maybe you should run more merchant now to enable you to capture that earlier to help counterbalance long-term rates if they end up being perniciously high, which is something I know on the team's mind and on our mind as well.
The other counterbalance is that on the development side of the business, we have seen for a long time that you are able to capture cost increases on the development side of the business through increasing PPA prices, particularly cost increases that are pervasive across the industry. So everybody is facing a particular change. Interest rates, in this case, is actually one of the biggest development costs, and so that tends to end up in PPA prices. So while it hits your back book, your opco, certainly before merchant curves come up on the development side of the business, you should not see that impact at all. This is a feature for the CDC valuation as well as you look at it.
But because they are developing at, as you saw on my AI returns slide, at higher returns on equity, there is a lot more headroom for that development side of the business to counteract this thing. But it is definitely something to be aware of. We do not know whether long-term rates will be higher for longer either. In some ways, in that scenario, the kind of AI end of the portfolio will be going great guns, but it will all need power. So somebody will need to figure out how to make that kind of work. All that contracted capacity that I talked about looks like this if you roll forward a couple of three years. So in the middle there, you have our updated guidance for FY 2027.
If you take into account CDC's contracted earnings that the team are working hard to deliver the data centers in order to make that real, then you take account of what Longroad is projecting will be their kind of in-year returns in 2029. You can see that CDC and Longroad on its own, our share of their EBITDAF will be as big as everything that we are generating this year. It is a high visibility to a very material increase in the in-year EBITDA of the group, as big as it is today, just from those two businesses based on what they have contracted today.
This is an update of the funding slide we showed last year, and we are now showing the impact of divestments and also reinvestments in the period, showing our funding flexibility continues to be good.
Obviously there will be future divestments in there and we continue to enjoy the benefits of our credit rating to enable us to deepen the funding that is available as well for what is ahead. We will continue to do that. Okay. So what does that mean for my pillars, Jason? A little bit, I think. This is kind of a reversion of it. You remember pillar one was the cash flow generators. I think we are thinking more about that as having sufficient cash generation across the business. This represents that a CDC is a growth platform, but also has the ability to pay more cash flow through to the group to meet the needs that I talked about before. So sufficient cash generation.
Just really a subtle change, but I think the lines were blurring between one and two, so we may as well address that to keep this current. Biggest change really is in pillar two, where we see scaled growth engines and their adjacent options. You should think about Longroad, for example, then you have adjacent opportunities in geographic diversification, so the smaller businesses that are generating long-term options. But then there are also these options that we talked about up in my advantaged position slide, where power and data centers can work together, so a joint venture between a CDC and a Contact. There are lots of those things.
That whole list that we put up on the screen are the types of things that we think are worth spending our time on, leveraging our advanced position to get into versus going and looking for too hard pillar three just now. Pillar three is still there. It is our future options in emerging sectors. Today, I would say that is probably the healthcare idea.
That definitely was the way we thought about it going in. Obviously, one of those businesses is in a sales process now. We also have Clear Vision, which we have not talked about before, which is a venture capital investment Infratil made a number of years ago now. The idea is that it gives us line of sight onto new technological innovations that are difficult as an infrastructure manager to get a hold of. It is happening in the venture capital space.
I think that investment and maybe others like it are becoming more and more relevant as we try and figure out what is beyond this kind of next frontier of AI. What are the new business models that actually we should be considering as future infrastructure? Not saying I am going to do it, but I think it is an important differentiator of Infratil that we are able to do, participate in that venture capital sector, this one happens to be in the U.S., and try and look ahead to what could sit well alongside a portfolio founded on power, data centers, and the adjacent options around them. Everything else stays the same. We still are fundamentally thematically driven, long-term societal needs that we want to be able to deliver as Infratil and Morrison team.
It has to have infrastructure characteristics, so long-term contracted cash flows are going to be a big part of the portfolio forever. We still need to have strong capital discipline. So although I am talking about longer-dated options in our offshore renewables assets, they do not get a free pass. It is not a free ride to create options for 2040, 2050. We have to do our jobs to allocate the right amount of capital to the right prospects. Not all of them will come off, but we are conscious that that is very important, and we have not changed our target return, nor our desire to continue to beat the heck out of it as best we can for you as investors. Let me finish.
One of the key messages I hope you have taken away today, we still feel the portfolio is incredibly well-positioned to build out around an incredibly strong core in the AI infrastructure space.
Got a nice upgrade from CDC team today. I will let them talk to that. We have good funding capacity now and a pathway to continue to improve that as the reinvestment opportunities emerge. I would really say we are going to be focused around adjacencies a lot. We are going to continue to scan for that third sector. Could be a big scale one we can transform, could be some of the exciting ideas Will are talking about. But we will be very careful on that, I think, as a team while we finish the job in and around the core sectors. But it will happen maybe longer term. We are super excited.
I bet the world has changed as I have stood up here, or it will change tomorrow night. But who knew? What a fantastic time to be alive and what a fascinating space for us all to be involved in. Thank you for coming. Thank you for your attention. I hope you have an interesting day, and I might see you right at the end of it. Thanks very much.
Morning, everyone. I want to give you a quick update on the Morrison business. We are a private business, not a listed business, so there is less known about us. But the relationship between us and Infratil is so critical for both of us, we wanted to give an update on what we have been up to. The picture that will pop up here in a moment is a great way to tell the Infratil story and the Morrison story in one page. This is the Tararua One wind farm with the Infratil board standing in front of it in the late 1990s. Anyone who has a long memory will know David Newman was tall, but he was not that tall.
That tall.
There is a little bit of a story here about scale. Also obviously about board diversity. I think we have made some progress on the Infratil board. But to go back to scale, when this wind farm was completed in the late 1990s, it was the largest wind farm in the Southern Hemisphere. Those are Vestas V47s for the tech geeks out there.
Those are massive 660 kilowatt equipment, and there were 23 of them. So it is about 30 megawatts there. Our first largest wind farm in the Southern Hemisphere at the time. You will hear about Amargosa today that the Longroad team are working on. That is 90 to 100 times the size of this. But obviously some time has passed, and if you think about that sums up the Infratil Morrison model, right? We try to be early, be ahead of the market.
When you find a great space, you build advantage in it, and when you find a great team, you back them, and then you grow scale. Over time you get bigger and you diversify, not just the gender mix of your board, but also the set of options that you can bring into a portfolio. So keep that in mind as we go through a few slides about our business.
This is a regular slide I have put up the last couple of years just to give you an update on a few key numbers. Most important is the one Alison mentioned earlier, the track record. That is still 18% per annum net returns for now 32 years. So that is ultimately all that matters, and everything else we talk about is just the ingredients to that. In terms of Morrison’s business, just give you a sense of our growth.
We are about NZD 60 billion of AUM now, so that has grown 25% in the last year. To be honest, we are a bit flattered by the weakness of the New Zealand dollar. Our assets are a lot in U.S. dollars, AUD, EUR these days. But yeah, good growth across the portfolio. The team today, I think on the slide we have 225, probably if we did a headcount, so that would be 230. I think when we put up this slide last year, we were about 215. So net increase, 10 to 15 people. But underneath that, there is actually something really interesting going on, which links back to Jase's points about AI, really. Ultimately, I am sure you are seeing this in your own businesses, the ability to analyze publicly available data is rapidly becoming a commodity.
Where real value accrues is secrets, private information, wisdom, the ability to make judgment calls about that private information, and influence, the ability to actually make real-world change in businesses. What that means for our team is much more focus on doing what Lloyd Morrison and I started, like grab really senior operators from industry, do not just fill the business with finance jocks, and use those people to fundamentally support, grow, transform the businesses that you invest in. We will talk a bit later about a few of the folks that have joined. Other important number on that page is one about alignment. Morrison on its balance sheet owns over NZD 500 million of Infratil equity as of a few weeks ago.
Not sure what that number is today with the movement of the share price, but it will be back there soon if it has dropped below. That is just what our balance sheet owns. Obviously, you see the disclosures from the Morrison family, from Jase, from me. You do not see them from everyone in the team, but a lot of us also invest directly.
That is really important because that performance number is driven by having an owner's mindset, and so us always thinking of every Infratil dollar as our own. Importantly, that applies to Morrison as well. Almost all of our staff and their families have their wealth and their future generations of family wealth tied up in Morrison and Infratil equity. It really does make all the difference when you are asking people to go the extra mile and really think long term about the business.
If you do the math, it is on average just through our Morrison balance sheet ownership, over NZD 2 million of equity per employee, and then you add to that the direct holdings of the team. That owner's mindset is super important. I know there has been a few questions about the partnership we announced a little while ago with Sumi. I do want to touch on that, because obviously that is the first time we have stepped away from this model of the Lloyd Morrison Trust, the founders and the team being the only shareholders in Morrison. First of all, who is Sumitomo Mitsui Trust Bank? There are a lot of different Sumitomos and Mitsuis out there in the world, so we should be clear about who we are talking about. They are the biggest trust bank in Japan.
They are the number one gatekeeper to the Japanese corporate pension fund market. What is the partnership? They have appointed Morrison as their preferred global manager for infrastructure. Why did they do it? Why did we do it? Actually, it is the same story for both of us. Japan has gone through 3 decades of deflation and negative real rates, and all of a sudden it has flipped to the other side, so capital is being mobilized. By the way, it is the second biggest pension market in the world. So a massive capital pool is being mobilized towards higher returning, more actively managed assets, but also with downside protection and inflation protection infrastructure. So really interesting for them, really interesting for us. I think really interesting for Infratil as well.
If you think you now have preferred access with the best gatekeeper to this giant capital pool, that should over time increase our access to capital for Infratil and for all the Morrison clients. Also, there is an amazing thing going on in Japan right now with positive real rates. Balance sheets of the corporates are under pressure to get efficient, so you are seeing a lot of corporate carve-out activity, embedded infrastructure assets being pulled out. The only way really to access that is to have a really trusted relationship with those organizations, and that is what Sumitomo Mitsui Trust Bank has, plus transformation capability, infrastructure investment capability, which is what Morrison has. So do not expect overnight results. That is not the nature of Japan. Also for Infratil, obviously, we have an incredibly high-performance target.
Even in a rising rate environment, not every deal we see in Japan is going to be right for Infratil, but I do think it is really interesting. Most importantly, there is no change to the control of Morrison, and in fact, as part of this, we have locked in, I would say in perpetuity, certainly as long as any of us are around the idea that the Lloyd Morrison Trust, left behind by our founder plus the team, will have positive control of Morrison.
But we will have one of the very senior execs of Sumitomo Mitsui Trust Bank join our board, which will be great just to keep driving the strategic benefits and alignment. So happy to answer any questions people have in the break about that, but I did want to share it. Last point I will make before I hand over to Will Smales, it is about people.
Where do we invest as we get a stronger balance sheet, a bigger business? We invest it primarily in talent and in investing in our own products and being glad recipients when the Infratil board decide to pay us performance fees in scrip, having a balance sheet that can happily accept that. So on the talent side, I will touch on a few important additions since we spoke last year. Brendan Scollans has joined us as head of North America, the role previously held by Perry Offit. Perry has moved to a different role, still in New York, running our global capital formation.
Brendan, you will see from his bio, has been global head of digital infrastructure at one of our global competitors, and previously head of U.S. TMT at another competitor. So great track record, great ability to invest and grow, particularly our digital infrastructure business in the U.S.
Gianfranco Saladino, classic European Italian, lives in Switzerland, educated in Germany, and previously ran the European value add strategy for another of our competitors. Really adding more of that language breadth capability. Todd Wynn actually came with great references from our friends at Longroad, a guy who has been a chief executive at another major U.S. renewable developer, adding to our operating capability. Think of him like, you all know Dion Campbell, most of you know Dion Campbell, like getting the Dion Campbell of the U.S. to support us there. I will not talk about all of the non-exec appointments, but we keep trying to add more perspective and insight around the boards of our portfolio companies, but also around Morrison.
The one I will mention is Kurt, because those in Sydney, Kurt will be down here next week, and if you would like to spend time with him, let us know. Kurt is the foundation member of our Morrison Advisory Board. He was back in the Obama era, if anyone remembers those days. He was Obama's Asia czar, and then in the last White House, he was Deputy Secretary of State, responsible for the Indo-Pacific region.
Right now, whether you like it or not, geopolitics is about the most important thing in the world of investment. So fantastic to get his perspectives. I will not talk about everyone on these pages, but you will have the slides available. Lastly, to talk about a few Morrison people who are in the room today who are not speaking, but so you get a chance to say hello to them.
You do not need to say hello to Philippa Harford. You probably remember her well, former CFO of Infratil. The reason I want to talk about Philippa is she is an operating partner at Morrison, sits on some really important boards for Infratil. Great example of the magic of this model. In a normal corporate, the number of really senior execs you can have is constrained by the org structure. What we can do for Infratil is hold on to amazing talent and apply them to the businesses where you need them.
We can have kind of a surfeit of capability and senior executive talent. So great to have Philippa around, and please grab her in the breaks if you would like to get an update on the businesses she is involved with. Grace is an unfamiliar face. I might ask Grace to stand up, if you do not mind, just so everyone sees you.
She is wearing a bright red jacket, so everyone can find her. Grace this year has been with us for a number of years, but this year became our head of Australia and New Zealand as Peter Coleman transferred up to Europe. Grace is a very long-term infrastructure investor and a wonderful person to know. She runs our investment team across Australia and New Zealand. Lastly, Clayton, who a lot of you will know. I will finish on Clayton because I started on the picture of Tararua Wind Farm. Clayton is kind of the bookend of that story. Clayton's career started out in Trustpower, the folks who built that first giant wind farm. When we created Tilt out of Trustpower, Clayton moved to Melbourne and became the chief development officer.
When Tilt was prised out of our hands by a great offer, we managed to hold onto Clayton and Dion within Morrison. He then went into Manawa with a really important job to do transforming that business and then driving the sale of that through to Contact. If you think about that long span of our involvement, like 1994 in Trustpower, Clayton wasn't quite involved right back then, but all the way through. That shows the way we try and transform these businesses build. When you hear about what we're doing between Longroad and CDC, I think that gives you another example of how we can step in, but also goes back to Lloyd Morrison, who was on that first picture.
One of the first things Lloyd said to me is, "I'm a magpie for talent." When you meet an amazing executive, you hold onto them, and you work out how you can apply them, and you work out how you can give them an incredible career path they couldn't have in an ordinary corporate working around the world on your portfolio. Grab these guys on the break.
I'm going to hand over to Will, who will talk a little bit more about investment activity. You can tell Will's been living in the U.S. by the size of his coffee. Good morning, everyone, or ohayou gozaimasu, as we now say at Morrison. I'm Will Smales, and yes, the clue was in the coffee. That's a large McDonald's regular coffee. The line at Starbucks was too long, so I went for second best.
I'm the Chief Investment Officer at Morrison. For the last five years, have been based in New York, now relocated back to Sydney, but obviously spending a lot of time across Morrison's business globally. By the way, I'm going to speak quickly today, so keep up. I'm going to start with the purpose. To invest wisely in ideas that matter, and that's Lloyd Morrison's founder. To invest wisely is hopefully intuitive for all of us in the room, but what that is a very well-developed, rigorous investment process where we say no to almost everything we see, and I'll talk about later when I represent some of those statistics that we review on behalf of Infratil. But importantly, on that purpose, what are the ideas that matter?
The ideas that matter, remember, are the businesses and assets that have infrastructure characteristics that come together against the investment themes that we're prosecuting and importantly, underpin the wealth and prosperity of society into the future. That purpose hasn't changed as Morrison and Infratil has scaled.
What has changed is the capability Paul was referring to apply against that purpose. It's important to work out, what does investing wisely and creating value for Infratil mean in the current environment? We are against this incredible sort of AI transformation backdrop. I was in Texas the week before last in Austin at one of the largest data center conferences, and my experience of the day sort of sums up where we are in the world. As I walked into the Marriott in downtown Austin, protesters were barricading the entrance to the conference center.
They were videoing every participant entering the conference center. On that same day in Austin, there were the leaders of some of the largest data center businesses in the U.S., the QTSs of this world, in front of the Texas Senate trying to explain their obligations on social responsibility, social license. At the same time that day, I had ridden in a robotaxi to the hotel, looking out the door at cyber cabs starting to roll out on the streets of Austin ahead of their launch in a couple of days' time. That world sort of explains that in Texas, which is becoming one of the largest data center markets, certainly in the U.S. and definitely the world, we have social license pushback. We have protests, yet we have autonomous physical AI rolling out in front of our eyes.
That is the context of what we are dealing with, and as investment professionals, we have to navigate through that, and investing wisely means having conviction in the trend. We have to be brave and bold and allocate capital and take risk. But we also need to be really careful and make sure we can pick through the hype and constantly check ourselves and say, "What could slow down this massive trend?" The final point, if you think about what are the trends underlying this, if you think about energy, we often with the AI hype, we forget that there was still this massive underlying trend around energy transformation.
So decarbonization of our electricity system and rise of electrification full stop. That trend is going on unabated, even underlying on where the AI trend is driving things. The same is true on compute.
We are hearing a lot with CDC that we are deploying GPUs on behalf of training and inference workloads. But the reality is we still had strong underlying cloud demand already happening within CDC, and that trend hasn't slowed either. It is important to pick through all of this.
The point Jason made is against this backdrop and these incredible trends, we have a massive head start. We have really scaled positions for Infratil with amazing management teams who are able to access this trend and do it with this information advantage. So when the Longroad team think about deploying capital against a new renewable development opportunity, or Greg is thinking about deploying capital against a new data center, the expertise, the information, the years of experience they have to decide to allocate capital at the right time is huge.
It is very hard from someone from the outside to compete against that, and that is why we are so advantaged in Infratil. By the way, through all that, we are still finding really, really exciting ideas which could become pillar three ideas for Infratil. Last year, I laid out the investment process for Morrison.
Again, this will not change. The important point to remember is the way we work at Morrison is it is an end-to-end process. The same people who are thinking about what is on the horizon, what investment should we make, are the ones then corralling those opportunities, executing those opportunities, participating in the asset management, then ensuring that company plays the right role in an Infratil portfolio or another portfolio of Morrison's. That sort of accountability end to end is really, really important.
I do want to talk about asset management. What is asset management? First and foremost, our objective is to make sure that we are supporting the development of highly capable, independent companies that are invested by Infratil. How do we do that? We do that through governance. We are applying best practice governance. We are making sure we are governing these entities properly.
We are making sure that we are working with the existing management teams to make sure they have got the right capability around the table. In the odd instance where we are talking about changing senior management, that is the role of the board and the shareholders. Importantly, we are sitting together as shareholders with management and the board, agreeing strategy and setting the risk profile for these entities. Primarily, that is what asset management is. We all know that the businesses that Infratil Morrison invest in are at different stages of development.
Importantly, Morrison is like a backstop. If you have got bursty demand, you are wanting to do a big refinancing, you are wanting to undertake a piece of strategic M&A, often those capabilities are not embedded in a portfolio company. Quite rightly so, because they may only be doing one of those transactions once every few years.
Those capabilities are within Morrison, and they are on call to all of our portfolio companies. That is a really important element of asset management. The other piece is what Paul mentioned, a deep bench of operating expertise. When we are trying to work out what our cybersecurity settings are in one portfolio company, that capability probably exists in Morrison or it exists in another portfolio company. Our job is to bring to bear the Morrison expertise, but also to coordinate across the portfolio. Some big numbers under there. Morrison today is managing 32 assets. Not all of those are Infratil assets, but the network, the experts, the capability is available to every Infratil asset. Here are a couple of really, really good examples, and you will hear from the One New Zealand team later.
When we made the investment in Vodafone New Zealand seven years ago, we were doing that on the basis of two very basic underwrites. One, we could separate out what was a subsidiary of Vodafone Group Plc and stand up an independent, proudly New Zealand telecommunications company. By investing in the network, we would create a leading platform in New Zealand to then drive a series of cost efficiencies in that business.
That is what we underwrote with the original investment, and I think management and the board and the Morrison people have done a wonderful job executing on that. We also knew that there were huge amounts of infrastructure sitting within One New Zealand, and you can see it there. First the towers are separated and sold. You will hear from the EonFibre folks identifying the fibre assets within One New Zealand.
We recently announced the RANCo idea. The idea that you can pool your radio access network resources, drastically reduce your future CapEx expenditure, and still deliver a better outcome for New Zealanders. That is a really good example of how relentless asset management drives those outcomes. Those outcomes are really, really hard for management to achieve, and they are really, really hard for owners and boards to achieve. The beauty of Morrison is we are able to sort of align our collective efforts and achieve some of these things. Jason referred a little bit to Anytime. So this is the idea that you could take capability out of Qscan and RHCNZ and bring it together and create a brand new teleradiology business. What does that mean? It is this idea that you can decouple the service of reading a scan.
You can make that highly automated, driven by AI, and you can scale it globally without needing the machines. Now, creating that business, and there are some folks in the room who actually made that happen, is really, really hard work. Day 1, you start with a really, really small business. So you are only going to expend that time, effort, and energy to carve out that business and create that platform if you believe in the long-term value that that can deliver to Infratil. So Morrison should be eking out in every corner of Infratil's portfolio value, applying sort of extra resource to creating that and then delivering that.
Then importantly, we will hear later how by connecting CDC with Contact, connecting Longroad and CDC together, we are hopefully able to unlock other opportunities in the portfolio. That collectively is a series of examples of asset management.
If you go back to that process slide, you remember right at the front is this idea that we are conducting research to identify what the new ideas that matter are. I showed this slide last year, and you would not be surprised that in the last 12 months, we have produced a huge amount of internal research, and that research is available to the Infratil board. We share that information with the Infratil board. Very importantly, what that research does is it helps us position and curate a pipeline early. This is a snapshot. By the way, there is way more actually than 200 opportunities over the last 12 months. But you can see the geographic diversity, you can see the increasing role of North America in our pipeline, you can see the split by sector.
Then importantly, on the right, we have listed some of the ideas that have been driven by research that are now starting to show up as opportunities in our pipeline. Some of the words there, autonomous mobility, sovereign AI, digital registries. These ideas are starting to fill up the pipeline, which means we are more likely to see them as high conviction investment opportunities. If you remember, I said we mostly say no to everything. This is a bit of a snapshot of what it looked like for Infratil over the last 12 months.
So we delved deeply into 20 opportunities for Infratil, and Anytime Radiology, which was an internally generated opportunity, was the only new investment opportunity. That is success. Actually, that is success. It is absolutely okay for us, for Infratil, to spend a whole year look at 200 opportunities and invest in nothing.
In fact, you should be really pleased we're doing that because we're looking at a huge selection of opportunities and being very selective. But we did invest in some great ideas this year at Morrison, and you can have both a great idea and the idea not being right for Infratil. As Jase mentioned, some of these ideas are very early stage, so not relevant or important to the context of a very large Infratil portfolio. I'll talk through them quickly, some of these ideas, and it'll give you a feel for the types of thematics that Morrison is thinking about. Polaris, this is a dedicated cold storage platform based out of Singapore. The trends we're backing there are food security. Singapore's an island nation that imports all of its food.
It's incredibly important for the government to have security of food supply that needs to be refrigerated in a lot of instances. So long-term, high-quality counterparties contracting that capacity, and then the ability to expand that platform in a highly geographically constrained market, coupled with the rising demand for specialist pharmaceutical products that need to be refrigerated.
That's a trend we had observed for a while. There are research papers we produced, and we found an opportunity, we executed on that opportunity. Felix, similar point. This is a greenfield platform in the U.S. In the U.S., as there is in Australia, there's incredible pressure on the pharmacy supply chain, particularly for pharmacists. Pharmacists are incredibly highly trained individuals. They train for six, seven years. In the U.S., a lot of them are spending time at the back of the pharmacy fulfilling scripts.
Those scripts can now be fully automated, so they can be done in a centralized facility by robots, counting pills, putting labels on bottles, putting them in a paper bag, and then dispatch them back to the pharmacy or to your home. In the U.S., some of the largest chains like Walgreens, CVS, have the scale and capital to build those automated facilities, but even they struggle in certain regions of the U.S. What we're introducing is a shared infrastructure model. Morrison will build and own these centralized fulfillment facilities and make them available to all participants in the market. Again, greenfield platform, really excited backing that physical AI automation trend. Orange, that's the JV in France. This is the recognition that nations must have GPU capacity on shore.
You need GPU capacity to process your sovereign workloads, but importantly, you need that capacity to be made available to create an AI ecosystem in your country. We partnered with the largest telco in the region to build sovereign AI capacity in that region. Again, a much smaller scale platform, and in case you hadn't noticed, Infratil has huge exposure to this trend, so it didn't need additional exposure in this format. Finally, it is a pretty exciting time. Jase mentioned it, but the way we're thinking about this current AI trend that's washing over the world is what are the enduring constraints in the system, and mostly they're physical constraints and infrastructure-like constraints. Jase talked about power. Connection to power, we think is the most enduring constraint.
We are spending a lot of time thinking about in our existing platforms or new platforms, how we do that. The other interesting thing, if you think about the second one, AI-driven productivity and transformation. We have learned a lot with the One NZ experience, and it will be great to hear the executives talk about that. But where you see Morrison or Infratil's name associated with complex operating businesses, this is what we are up to.
We are taking a view that AI can dramatically reduce operating complexity and basically shrink the business down to the core infrastructure, embedded infrastructure. And where you have large-scaled platforms with embedded infrastructure and a shrunk complexity, you get really great enduring investment thematics. Where you see our name associated with that is what we are trying to do. Finally, on the right, physical AI.
We have barely touched the applications of AI in the modern world. Yes, we have seen displacement of a few software engineers or increased productivity of a few software engineers. The next wave that is going to come through the economy is physical AI. And how we position to provide that will be critical. Morrison is already, if you think about that pipeline I showed you, thinking about many physical AI infrastructure investment opportunities. Thank you.
Thanks for that, Phil. We do have 5 minutes for questions, which can either cover the Morrison piece or if there is anything around Jason's introduction. We might hold the specific company pieces till when the MBC companies actually present because they are going to obviously be talking about things in more detail, so maybe at a capital allocation level.
Hey, Ben.
Hey. First question from me just on how you think about the portfolio in terms of-
Yeah.
-these data center assets, once they become mature or once they become developed, do you want to hold these for maturity or do you think the Infratil success over the years has been monetizing these assets? Is that going to be a story that will play out in the next five years for Infratil, do you think, or will you hold them?
It is definitely an option. Greg's talked about this before, and he may talk about it a bit today. It is easier said than done, finding the right way to do that for our customers, for the team in terms of operating them, and then obviously from a capital allocation efficiency perspective. But that is definitely a work stream that is going on, and that technology is being invented around us. I would repeat what I said before, is we do not need to invent it, but it is definitely something that we should keep an eye on. Just portfolio mix, exactly. Mm-hmm.
Yeah, Jason, it is Phil Campbell here from UBS.
Oh, hey Phil.
Just wanted to get a bit of a clarification just on the change to the pillar 1.
Yeah.
I assume the cash flow, especially cash flow generative, yeah, it would be quite good to get your views on what that actually means.
Well, we have what do we need a year, Andy? Like NZD 400 million to keep the lights on. So basically, we need to crib across the portfolio a steady stream of NZD 400 million. One NZ is providing nearly half of that, get a good chunk out of the airport.
We do get a bit out of Contact Energy, obviously, which is in the energy spot. Then, the balance really from across the portfolio, but clearly line of sight to CDC doing that. Is that kind of what you are asking? I only put those 2 there because that is kind of their primary role, but it comes across the portfolio, which is kind of why I made it a principle rather than there are only these 2 ones in the box. We are cribbing it from across the platform, yeah. Here we go.
Hi, it's Owen Birrell from RBC.
Yes.
I just wanted to touch on something that kind of made a small entry into one of those slides.
Yeah
Around the NZD 1 billion worth of potential divestments.
Yeah.
Because we talked about all the assets you have.
Yep
and it seems like it appeared on those slides you're quite wedded to them.
Just wanted to get a sense as to how you guys are thinking about that NZD 1 billion
Yeah
of divestments. Is it existing assets or is it spinoffs of parts of existing assets
Yep
such as the tower sale?
Yep
going forward? How should we think about that?
Yeah, thank you for asking that question because I didn't mean to put everything there in the portfolio forever. I'm just trying to give you a sense of, excuse me, where they sit. So there will be elements of spin-outs of existing assets that have been talked about as well. We're obviously midstream on, I guess, roughly half our healthcare portfolio. Probably our view hasn't changed on that in terms of future optionality. It'll be a bit of both. Nothing has really changed there. I think what I was more trying to signal is there's a bit more tolerance for smaller scale, longer-dated options in and around our core pillars.
We were reflecting, asking Galileo to get up to $3 billion or EUR 3 billion in three to five years in that market is unrealistic. You sort of say, "Well, do I want to be completely out of the European market with options in the part of the AI value chain that seems to be the most scarce and enduring forever? Or would you go out and then buy back in?" It's like, hmm, we should think about that a little bit harder, I think. More tolerance in and around the core pillars, I think, for starting smaller scale. Even some of the adjacencies, right, a transmission or something. Again, if it's related, it's not dragging too much in and around our core pillars. Maybe you are right to start smaller and build up in and around those core pillars.
I think if you're in a future option, a genuine future option is probably a little bit less flexibility.
But in terms of, yeah. No worries. Grant, did you-
Hi, Jason. Grant from Jarden. Question around CDC and Contact.
Yeah.
CDC's strength is obviously building data centers.
And with that comes the contracted demand that we will focus on.
Why would you dilute that by moving into something where the power supplier gets a share of that IP?
It's a good question. You can ask the CDC team their version of that. Maybe to zoom out just from that particular idea. I think there would need to be clear advantages for the data center operator to do that. They will be things like speed to power that is consistent with the kind of social license requirements that we've talked about as well. I think that'll be case by case, looking at situations where you can get there faster at a scale than maybe you could otherwise with a strong social license story. So individual situations. Then really it just comes up to the dynamic at the time as well.
I think it's also within those JVs, right, Jason, where within those JVs, we're also thinking about how each party then gets value from that. Depending on what the party brings, probably determine what they will receive out of that JV. So, as we go through those. Those set ups, we're thinking very carefully about who we bring.
Carefully weigh that you're getting a fair share of.
Yeah
the benefits of that particular site, I think, yeah.
While demand is running so far ahead of what you can supply at, you will keep looking at those sort of opportunities.
I think we should. One of my reflections on that AI value chain, when you come up with the returns, is that we are selling up how cheap generally. If that is really the key gateway, are they not similar to further up the chain, particularly when you go all the way to the right and you see the crazy returns you get from using that power. I think that will shift naturally over time. But at this point, yeah, why shouldn't power earn more in this environment? I do not know. Yeah.
Great.
Thanks for the presentation, guys. Stephen Hudson from Macquarie. Just two from me, just following on from that question. Is it grid connection or is it green electrons that is sort of really the enabler now for data centers? I was sort of interested in all three of your touch points there.
Look, I think it is going to be both, as social license stuff comes to the fore everywhere. Clearly, there is an issue of transmission that probably creates the obvious constraint, physical constraint. But unless you come with a good power story, which I think was probably the advantage of the Contact JV in particular, it is not a sustainable business model, I think.
No, I agree with that. I think the grid connection is a somewhat binary outcome. You must have a grid connection to be able to stand up your data center, apart from the odd behind the meter stuff. I think the green elections thing then becomes what is the cost of your electricity and how much are you being asked to contribute to that cost as a supplier. So that for me is more a variable, whereas the actual connection to the grid is a hard constraint that you have got to overcome. So in that sense, the grid connection in my mind comes first and is probably more important. Having said that, we still need to secure green electrons as well.
So your forefathers had it right in the '80s, really, given the sort of DNA of the company.
Yeah, would have been. Yeah, would have been.
Yeah.
And just interested in your comments on infrastructure. I think, Jason, you sort of said-
Sizing it.
It is downside protection and inflation protection. Of all of your assets, which actually provides the inflation protection best? I was a little bit surprised with what you said about Longroad, because presumably- Sorry, I am answering my own question here, but -
Presumably some of those PPAs are indexed in the same way that sort of Contact JV-
Not indexation anymore. You have CPI linkage and a bunch of or escalators in data center contracts. It is all in the val already, right? Yeah, change the assumption, you will get a different outcome still. The airport is the best inflation-protected asset. It is not very hard to figure that out, the way the regulated asset base works. So that is easy. One is actually not bad.
Yeah
because you can reprice more often, and as long as your market structure keeps right in the consumer side of mobile in New Zealand, that market structure is good. So yeah, in my head, I go airport one and then data centers with limits. The other thing to think about is half of the vals of the data centers and the energy businesses is development, and you can recapture all that inflation in your new developments as well. So yeah, that is the answer. Is there something from you?
Thank you. Suraj Nebhani from Citigroup. Just one question for you, Jason. You mentioned in the presentation around interest rates impact on Longroad.
Yes.
My understanding is NAV growth is the big part of the story for Infratil as you grow the value of the businesses. How do you think about that with higher interest rates? I guess a follow-on to that is.
Yeah
Why are the return requirements not changing if your funding costs are going up?
Yeah. I think part of the answer, maybe to answer your last question first, is our hurdle is pretty high anyway. It is 12, pretty clear air between that and long-term interest rates. We are in private markets, really. You cannot just change on the fly. There is value, I think, in just keeping it clear with your teams what your absolute return requirement is. We have had this out of the back of COVID when interest rates were flying all over the place as well, inflation expectations are flying all over the place. I think in terms of the NAV with real interest rates, you are just going to have this headwind for your existing long-dated contracted assets, which is half of data centers, half of Longroad.
Then you are going to have to catch up through the price at which you pay, at which you sell your new developments, which is basically the other half of the business. I think of it as actually reasonably balanced for the defensive and the aggressive at the end of the day.
It is kind of interesting, right? If AI slows down, then I think there is every reason to think real rates are pretty low, so 50% of your valuation is not going to now, that could easily become 75, right? But then if it charges hard, then actually 40% of the val in development, because if AI charges hard, that pushes long-term growth higher, which means long-term real rates are higher. This is kind of basic thesis. If that is true, then the debt co looks pretty undervalued to me.
Yeah
Because your debt co will run harder. For all the reasons I have talked about with these physical constraints, you will probably capture more. I think it is kind of difficult, right, with a snapshot, but either way, you should have these counterbalancing forces in the val, that as long as we continue to execute well, as long as we are right on the physical constraints, as long as demand continues, which is all the things we have talked about today, there is every reason to find your way back to kind of the central course that we are.
And we have been running against this headwind for a while now. I mentioned sort of the 30-plus assets across Morrison. Almost all of them are doing regular independent valuations. We have a very good view of how valuers are thinking about risk-free rates. Obviously, this trend of rising rates, you know
Has been a headwind for a while in all of our independent valuations. I guess the point Jason and I and all the-
You work hard-
-worries are what? We are sitting like, when does this let up?
Yeah.
When do we actually see valuations, from a risk-free rate position stabilizing? I think we're all on edge at the moment in terms of what we're seeing in long-term Treasuries.
If you think about what's really going on there, right? It is we've gone from a world where there was too much capital chasing not enough assets to one where actually the world's quite capital short, and increasingly so the way governments are going plus AI CapEx boom. I think when you hear from the Longroad guys when they talk about their recent acquisitions, that shows the advantage of having-
operating capability and flexibility to act. Across the Morrison business, we've probably seen in that renewables space, kind of the attractiveness of the opportunity set in U.S. renewables just kind of go through the roof as rates have risen, and you've had building capital constraints. I think that's the other half of it, right?
Yeah.
How can you act with speed and compete-
Yeah, from your platforms.
-when the world becomes capital short.
Mm. Okay.
Are we on, Fletch?
We might break there, but, thank you, guys. We are going to have a half an hour break and then start back just after quarter 2. Thank you.
You are welcome.
While we have the Longroad team come up, I was reflecting in our recent investor relations travels, we have been in the U.S. and Asia over the last, probably four to six weeks. What has been really interesting is, historically, and it is just the nature of business and where the growth has been, there is a lot of talk about data centers.
We probably spend 90% of our time talking about the data center industry and obviously CDC. What is interesting in the last sort of two to three months is that has now been balanced out by people wanting to talk about power and the access to power. That is regionally and it is globally and obviously, we are pretty excited about the business, which is not a new business. It is 10 years old next month, I think October the 4th.
These guys are sort of the quiet achievers within the portfolio, and they have got a probably incredible position in the U.S. Welcome them. I do want to acknowledge Michael Alvarez, who is sitting at the back there. We would normally have him up here. Somehow he has managed to get out of it this year. Please see him at lunchtime or any time during a break, because he is a wealth of knowledge and experience. Thanks, Paul.
Okay. Thanks, Fletch. Good morning, everybody. We're going to try not to be quiet this morning, Mark. We've been quiet achievers, but we're going to hopefully make some noise today. I do want to introduce my colleagues, Pete Keel and Charles Spiliotis, as well as, Fletch just mentioned Michael, and then Jeremy Law is here, too. He's our new COO. I'll talk a little bit about that in a second. Thank you, Alison. Thank you, Infratil board. Thank you, Jason, for inviting us back for another tour of duty here in Sydney. It's beautiful to be here. We're really happy with the progress we've made over the last 12 months. We're going to go into some details. We actually have a pretty solid slide deck for you all to go through.
We'll give you some color on what's happening in the market, what's happening with some of the challenges and opportunities in the U.S. We'll talk about our business, our outlook for, really kind of to the end of the decade, to the end of 2029. We are turning 10 years old, believe it or not, in a couple of weeks. I look around the room, and we weren't invited to this nine years ago.
We didn't make the cut. But we've made a lot of progress. We're super supportive and very happy with Infratil as our investor. We've also had great support from NZ Super Fund and, more recently from MEAG. So we're really, really excited about the next 10 years. But it's given us a little bit of time to reflect on where the business is going, and I'll talk a little bit about that.
But we call it NextGen, the NextGen of Longroad. So 10 years down, 10 years to go, a long way to go. But I think the general theme of NextGen is what you think worked well for you over the last 10 years is probably not going to work that well for you for the next. Sorry, for the previous 10 years is not going to work that well for the next 10 years. We've been trying to embrace that, reinvent ourself with a little bit of a management team shakeup. And we're trying to make sure that Longroad is ready and built for the future. Again, let me kick in and let you know what's happening in our business. So market demand, Jason talked about it this morning. Charles has some slides.
But a really generational opportunity in the power business, in the renewable business in the U.S. Think about anywhere from 150 to 190 gigawatts of installed capacity between 2027 and 2029. So that's an unbelievable market dynamic that we haven't had since we started Longroad. We are really happy to update our guidance. I'm going to talk a little bit about that on the following couple pages. We're increasing our gigawatt target 25% and our EBITDA target quite a bit more. In order to do this growth, to hit this growth curve, we need capital. We are really happy that Infratil invested in our internal round in March, that they reported out on that earlier this year. And that capital has us funded into 2027.
Then we're also starting an external capital raise in a couple of weeks, where we're going to try and find the next block of capital that's going to fund us to the end of the decade. Maybe roughly $750 million to $1 billion of capital. The renewable business remains challenging. There's all sorts of issues. I'd say the top three friction points right now in our business, regulatory, that has a lot to do with tariffs. Executive orders that have not a lot of definition. What do we do with this? We don't know what this means. So there's definitely some regulatory friction. The cost curve has gone up everywhere, right? Inflation is real. Panels are more expensive. Batteries are more expensive. EPC costs are more expensive. Interest rates, we'll talk about. So the whole cost structure is going up.
But luckily, because of this market demand, we feel like we're getting a fair bit of relief from the PPA counterparties. The one other thing that Charles will talk about a little bit is interconnection queues that are pretty remarkably long. So that's another point of friction. One of the places where we made a lot of progress over the last 12 months is on our digital infrastructure strategy.
It's a little bit adjacent business from our core business of renewable power. But there's a lot of overlap, and I don't want to steal Charles' thunder, but we have roughly 10 gigawatts of data center opportunities, that are adjacent plays with our existing portfolio. Then I touched on this management team. This is the NextGen thing. If you look at that logo there, how creative is that? Right? See the 10 Longroad? Right, 10.
So that's just what a bunch of finance and engineer guys can come up with. But you can see a decade down the road. So this is this NextGen concept and ability to refresh our management team. We spent the last 18 months thinking about succession, and at the request of our very diligent board of directors, and we put a lot of effort into it and a lot of thought into it. And we made an announcement in July, about kind of the next generation of leadership. So we're really happy with where it came out, with a ton of work that everybody put into it. Okay.
So on the guidance, if you look at, so last year, the gray bar is what we told you last year, and we didn't tell you anything about 2029 last year because we didn't know, but assume that we were going up 1.5 gigawatts a year, so 10 to 11.5. You can see that the number went from, if you imply that 11.5 gigawatts in 2029, our guidance is now over 14 gigawatts. Last year, we told you at the end of 2025, we were going to be at 5.5 gigawatts. We came in a little bit below that at 5.3. The project that we thought we were going to close in December actually ended up closing in March. So it just slipped a little bit. We're on track for hitting 7 gigawatts by the end of this year.
Just a little bit under that. We have had a few assets that have kind of swapped in and out. The blue bar is what we have. We have a fair bit of work to do for the rest of the year, but we expect to basically hit our targets.
One of the things that I am going to talk about a little bit is, and I know Paul talked about it, the Amargosa acquisition. That is really the main driver for us revising our guidance upward. We used to talk about doing 1.5 gigawatts a year. We are now averaging 2.5 gigawatts a year. That might not sound like a lot to you guys, but it is a big lift. We are really happy with where the guidance is at this point. On the run rate EBITDA, same story. Number is a little bit different.
You can see again, we told you last year at the end of 2025, we would be at $380 million. We are actually at $410 million. We are going to outperform our expectation a little bit this year, mostly due to the asset mix. Then you can see we get to $1.2 billion of EBITDA by the end of 2029. That is a pretty sizable business.
This is kind of, again, I am trying to get everybody to think about NextGen. We are not this startup little company anymore. This is a really big company. Pete is going to talk about what does that mean when we are at $1.2 billion of EBITDA, what does that mean for our financial stability future? What kind of options do we have at that point? Okay, what have we been up to? We put 2 projects online so far this year.
Like I said, we have got another 1.1 gigawatts to go. I mentioned the Infratil NZD 300 million equity raise. We have a 36-gigawatt platform that does not include the data center business.
That is just our core business of wind, solar, and storage. One of the things that has become remarkable, if you asked me 10 years ago, how much do you think you have invested in your pipeline? I would probably say, I do not know, $250 million. We have $1 billion invested in our go-forward pipeline. That just goes to the scale. Pete is going to talk about scale, making sure the business has scale, and having access to our operating portfolio gives us the ability to finance some of that a lot more efficiently. But we have got $1 billion of forward of development capital working on deals that are out into 2032, 2033.
One of the big questions last year was OB3, and actually came out in a very constructive way for us. One of the things that came out of that was the rules around tax credit qualifying or safe harboring our projects. You can see we have got 15 gigawatts of projects safe harbored.
You ask, okay, if you are only trying to go from 7 to 14, that is 7 gigawatts, why do you have 15 gigawatts? The reason is we are not exactly sure if projects are going to fall into 2029 or 2030. Therefore, you have to almost over-procure to make sure that you have got coverage in the instance that it slips from 2029 into 2030. Then, the digital infrastructure pipeline, I do not want to, again, steal Charles' thunder, but we have been spending a lot of time on it. We have been learning a lot about the data center business.
These two businesses, the power business and the data center business, are inextricably linked, there's no question about that. I'll talk a little bit more about the Amargosa acquisition on the following pages. As you can see, it's a big one, 2.8 gigawatts solar and storage in Nevada. I'll give you some more details in a second. Here's some pictures of projects that we put into service. Thousand Miles, this is in Texas, but it's not an ERCOT, it's an SPP. SunPond is the fourth. Some of you have heard of SunStreams. Some of you have been to SunStreams. This is the fourth installment of our SunStream strategy, so it's been a great way to grow off of that initial investment. This is what we've done so far this year. Fremont is a project.
This is the one that slipped from 2025 into 2026. That's a nice little deal in Utah with UAMPS, which is a conglomeration of municipalities in Utah. Milford Transmission. This is a project that we've— I'm sure you guys have heard of Fervo. Fervo is a geothermal company that actually has their first big power station right next to our wind farm in Utah.
We are actually selling them transmission services so that they can evacuate their geothermal power into markets west, so into California and other parts of WECC. This is a good example of playing offense. Again, we have a pretty sizable portfolio, 35 projects. One of the things we're trying to be a little bit more thoughtful about is how do we leverage our existing portfolio. You're going to see that in vivid color when we talk about the digital infrastructure strategy.
This is another great example of basically leveraging our existing assets on this transmission deal. Here's what we have left for the rest of 2026. Notch Peak is a project in Utah. Again, it's same kind of strategy as leveraging. It's actually at the northern terminus of this transmission asset I just mentioned. We actually acquired that project from another developer, then we've contracted with an existing relationship that we have, Southern California Public Power Authority. So we hope to get that deal closed pretty soon. Rosemary is a nice little deal in Fresno County, selling power to Ava, which is a community choice aggregator in the East Bay of San Francisco. Raven is a small deal. I would call this more of an experiment for us. It's an experiment that we thought about rolling up these small distributed assets.
I think financially, we did fine on this deal, but in terms of the roll-up strategy, it didn't really make a whole lot of sense for us, just given the scale, right? 31 megawatts, look, we're trying to do 2.5 gigawatts a year, you just can't get the bang for the buck on the human resources side. Milford Transmission ll . This is a further expansion of the deal I just mentioned with Fervo. Again, leveraging excess capacity that we have in a transmission line so that they can evacuate power to points west. Finally, Amargosa. Amargosa is just, if you're in Las Vegas, it's probably an hour drive to the northwest. It's very adjacent. It's kind of across the state line from Death Valley and Area 51, for those people who have that kind of interest. It's an interesting deal.
This deal was actually initially developed by NV Energy. NV Energy is the utility in Nevada, and it is a subsidiary of Berkshire Hathaway. They were developing it on the back of a pretty big transmission build-out that they are building, kind of a north-south link called Greenlink, and Amargosa is right in the middle of it.
It is a hugely important project for them, but from a regulatory point of view, they could not actually end up owning this project. They decided to put it out to market, to the developer market, and we participated in an RFP and were successful. There is still a lot of work to do with Amargosa. I think just a couple of points to make. Number one, what is driving the Nevada growth is data centers. Nevada, NV Energy receives something like 20 gigawatts of interest for new data centers.
From a siting point of view, Nevada has a lot of advantages, and you can see the demand. I think the system is only 15 or 16 gigawatts in Nevada, so they get more than their entire installed capacity in terms of interest in data centers. Having energy and renewable energy in this project is a hugely important factor for NV Energy.
The other interesting thing is that the project comes with a PPA, actually two PPAs, there are two phases. In the PPA, there is an indexation mechanism for the world without tax credits and a world with tax credits. Right now, the projects are expected to qualify for tax credits in 2029 and 2030. But if one of the phases slips, there is actually a mechanism pre-built into the PPA that provides a higher price so that we can actually build a project.
That just gives you a sense of how critical this project is to NV Energy, and it gives us also a lot of comfort that we are talking about pretty substantial development dollars at risk, and we want to make sure that at the end of the road, there is going to be something to actually build. This gives us the comfort to do it. We are really, really happy with it. I am going to turn it over to Charles to talk about the market. Thank you.
Thank you. Hi, everybody. We have been doing this a long time, the team here. Obviously, we have been at Longroad for 10 years, as Paul said. Before that, a company called First Wind. I think something that we are pretty excited about is what you see on this slide. For the balance of that time, renewable energy development in the U.S. has been a highly politicized investment, highly environmentally driven. If you were a proponent, that just meant that you are one of these Democrats wasting money, tax credits, all these sorts of things. I think that is a pretty tough characterization. That, nonetheless, was certainly evident as we went through our lives and you operate in different states and different regions that have very different opinions about what we are doing.
I think this slide captures a real significant shift that we are all extremely excited about, which is, while there is still politicization, as we see in Washington, it is a much different feeling when there is such significant demand for new energy in the U.S., driven, of course, by the digital build-out, on-shoring, manufacturing, electrification of vehicles, and everybody wants to ensure that we are ahead in the AI race and that we have the infrastructure required to retain the position that we have.
That is a pretty huge shift for what we are doing on a daily basis. We are out in a lot of rural communities. We are out in a lot of places where historically there has just been real challenges, and now it feels much different. There is just a real significant shift in the way that we talk about what we are doing.
Grid reliability, need to get the electrons online to support hyperscalers, support new manufacturing across the U.S., and that feels really satisfying, I think, for the people up here doing this, who feel like we are doing really important work. Here is some of the growth. I think the theme of several of these next slides is the U.S. is a big place. We are doing a lot. You see up there in terms of 3% annual electricity market growth, somewhere in the 2%-3% across basically all expert analyses on what we can expect, driven by some of the sectors that I mentioned before. The numbers today are actually relatively small in sort of installed base of data centers in the U.S. The growth is significant.
We talk all the time about hyperscalers, but you can see right here, about 78 gigawatts by 2030 of demand outside of the hyperscalers. It gives sort of a sense of the size of the opportunity. We are looking today at $40 billion-$50 billion per gigawatt of digital investment. So sort of the scale at which load is growing in the U.S. is pretty amazing when we consider basically we did this for the first 20 years with no load growth. We were almost entirely building new renewables in order to take coal offline, and it is a very different goal today. There we go. That is still happening. Coal is still not economic in the U.S., and never will be. There will not be a new coal plant built in the U.S.
Only the pace at which they are retired is really the only thing that folks in Washington and elsewhere, state-level administrations can control to a certain extent. But the inevitable is happening, and it is just how quick across the country those plants are retired and replaced by renewables. Here is the numbers on gigawatts.
Just to scale the market, I am going to slightly increase the numbers Jason said earlier, but about 60 gigawatts of new generation built in the U.S. last year, about 90% of that is renewables, so somewhere in the 55-gigawatt range. It is where all the money is going. It is where all the opportunity is if you want to build new infrastructure fast, at economic rates. I think the numbers here are pretty staggering. 70 gigawatts is a number that I do not think a lot of us would have believed a few years ago.
But incredible opportunity for us, the market dynamic all the way across the country. For now, a lot of it is in certain places we are trying to take advantage of the tax credit before 2030. Certain markets that certainly are focused on locking in lower prices for the grid, local grid or elsewhere. Then some markets that just have incredible load growth, and they need the power really at whatever price it requires, so long as it is fast.
Here is the challenge slide, that Paul alluded to before. So there is a lot of challenges in our business. It has always been highly volatile. Washington has always had a pretty big impact on what we are doing, in some way or other. Now we have tariffs. That is sort of a new one in terms of the way we have been planning. We had it since the first Trump administration.
To some extent, it has always been more manageable. The volatility is really the hard part. So we are trying to buy equipment, trying to build really long-lived assets. It is taking longer to acquire the equipment, as Jason talked about the supply chain challenges, things like that. A pretty big shift we have had in terms of what Jason was talking about earlier, in terms of locking in the revenue.
You still have all this volatility on the cost side. It has been a pretty big shift in terms of when we are actually locking in those power contracts. When we were doing this 10 years ago, you used to want to sign the PPA as quickly as possible, invest as little money as possible, sign the PPA, and then go invest aggressively to get projects built. That is sort of the opposite of what we are doing today.
We are out there investing in interconnection positions and land, trying to find the best sites that we possibly can, knowing that if we do that, and we develop the best projects, that there will be a revenue contract at the end, and we will be able to manage the cost side and be able to guarantee that that is a profitable project. So it is a very different dynamic for us.
We have been able to manage it, I think, well, with a lot of our OEM relationships. So for solar, we have gotten up here and talked about before, how we have about 3.5 gigawatts, under our frame agreement, still kind of fixed price, locked in, not subject to the tariff regime. That is a pretty big advantage we have today. There is a lot more domestic manufacturing in the U.S. than there was three or four years ago.
That has been a real change. That is on panel side, battery side, new cell manufacturing. There is a lot of opportunity for folks like us, and a lot of it is at the scale we are operating at. Very difficult if you are a small developer to go out there and lock in a lot of these costs at scale. Both because manufacturers want to talk to the end users, the ones who are going to own the asset for the long term, who can do really large volume.
That has been, I think, a real competitive advantage of ours. Build costs are going up. All this stuff is. Labor is a huge input cost. Construction competing for the same labor that the data center folks are. Lots of commodity price increase, tariffs of course. But the PPA market has been receptive.
Prices are still at a very competitive level. The point I just mentioned earlier. We are able to sort of lock in a lot of these issues before we sign that PPA. It has only moved one way over the last several years, but we have had pretty good success in terms of optimizing to provide the best value to ratepayers while still building economic projects.
The last one is the interconnection queue. It is one you hear about a lot. I think this is why we feel so positive about our development pipeline and positioning. We have invested in our interconnection queue for a long time. I think in many places in the U.S., seven to 10 years is not unusual nowadays. So we have a 35 gigawatt pipeline, which I will talk more about. That is a lot of investment that has happened over the last five to 10 years now.
In terms of trying to get us to a place where we have these positions, we have competitive advantage. We are always doing a ton of analysis around where we see new transmission investments happening, where we can position ourselves best. We have a really strong interconnection team. A lot of the tools that are available to us now, AI and others, can only help us create leverage for us to accelerate, do better analysis, find more opportunities. So it is expensive, takes a long time, but it is another reason that having a bigger platform is really advantageous. This is another, the U.S. is a big market, slide. So on the left hand here is kind of corporate PPAs first half of this year in the U.S. versus the rest of the world.
There is a lot of activity, and you can see in 2025, 66 gigawatts of corporate PPA signed, which is a pretty staggering number. Some of those are shorter dated, some of those are longer dated. But in terms of the continued demand, hyperscalers and non-hyperscalers. That trend has certainly not slowed, and it looks like 2026 is going to be similar, especially as certain folks are trying to lock in renewables while the tax credit is still available to them. Here is the lowest cost source of new generation slide, which is renewables is incredibly competitive. The speed point, I think we have already made. We can move faster than most of our competition. But despite a lot of the friction, and sort of tariff-driven skewing of the economics, it continues to be incredibly competitive product.
As we start to plan for post 2030, which is when the tax credit for solar and wind expires, and largely today expected not to be renewed, this is the unsubsidized slide that we really view as incredibly important. I just mentioned seven-10 years, right? We are making new investments today for 2035 projects, 2033 projects. We need to feel comfortable that we are competitive in those projects. Some of what you see here is some of the benefits of the technology improvements. The battery storage is probably the one with the most room to grow. On the solar side, we will see continued efficiency of the technology, but in terms of the shape of battery storage, it still has significant gains we think to be made in terms of competitiveness, cost, and long-term durability.
That solar PV plus storage utility is really where we are focused. We think it can be competitive with gas for the long term, which is really what we are competing against in most of these applications, right? Gas takes longer. It has gotten much more expensive. Cost of new gas plant has probably doubled over the last three or four years. We have seen inflation as well, as we have already mentioned. When you think about the speed factor, when you think about the lower risk nature of it, we are feeling really well positioned. This slide is showing the point that I already made, right? The right-hand side here, California, Arizona, and ERCOT. Markets where we are focused, we feel really good about our competitiveness. That lowest unsubsidized LCOE analysis is of course across the whole country.
Our pipeline is most heavily focused in the Western U.S., and in particular, the Southwestern U.S., which has the best solar resource in the country. It has the most available land. You can build at really large scale. A place like Arizona, which has, call it, the best solar resource in the U.S., has a battery storage tax credit, which lasts for another decade. You are getting a bit of a soft landing, as we say, with the battery storage credit extending for that time period. In Arizona, in California, a solar plus storage product will be incredibly priced competitive for a long time, we believe. The charts here on the right, I think illustrate it.
I already talked about the best outlook and why we feel really good about the declining cost curve there and technological investment and improvement that is happening. A lot of domestic manufacturing and domestic investment in the battery supply chain. In terms of the markets, west is where we are heavily focused. MISO and SPP are growth markets for us, where we see a ton of the data center demand, and relatively underbuilt, in terms of the wave of renewables. We continue to make investments there. Some of the projects that Paul Gaynor highlighted earlier. We see a lot of opportunity in those markets. This is getting back to the plan that we have here. In 2025, we are about 1.5 gigawatts. 2026 and 2027, similar numbers. Trying to illustrate here that we have a big pipeline.
We have a lot of coverage on our pipeline. Even though we say we have a 1.6 gigawatt goal for next year, we have 7 gigawatts of potential projects. Our business is highly volatile. Things generally take longer than you think they should, longer than you think they would. We need to have a lot of options in order to make sure that we can hit those targets.
You expect you have the best project, and you bid the PPA, and you miss. Next RFP is a year away. I think for us, it just goes back to why we feel really good about that NZD 1 billion that we spent investing in our pipeline and creating options all over the country to try to create real diversity and ensure that we can grow at the rates that we hope to grow at.
You see the step up that Paul mentioned in 2028 and 2029, that is on the back of the Amargosa project that Paul went through. Digital infrastructure. I am trying to remember if we sort of teased this at all last year. This is the first time. I am not sure. We are pretty excited about this opportunity. I am personally spending a lot of time on it. I think if you look around Longroad's portfolio and where we view sort of exciting opportunities, we have 6 gigawatts going quickly to 7 gigawatts here operating fleet. Which means that in those 7 gigawatts, we have been doing business in these places for a long time. We have relationships with landowners. We have relationships with the interconnecting utility.
In most of these places, we have growth already planned or we would like to expand or grow or find other ways to do more in these places. We have familiarity with the permitting regime and probably local relationships.
That is sort of the most obvious. It is like we have this operating fleet of 7 gigawatts today. It is going to be 14 gigawatts in three years, and there is a whole lot of options in there to leverage the work that we are already doing, and find ways to invest in digital infrastructure. It is really two ways that we can do that. There is probably 3 ways we can do that. There is obviously co-locating a campus with our renewables, which is certainly what we are doing in a lot of places. There is the opportunity to do smaller scale.
When Will is up there talking about the folks out in Austin who are out protesting, they are probably not going to protest at a solar site that is already 5,000 acres, and we are using the same footprint, and we are adding relatively modestly, with no permitting requirements and no real interconnection requirements.
Those are pretty, we think, interesting opportunities because of the speed and the relatively low friction of investing in smaller scale. We think the U.S., the opportunity set at the smaller scale is really interesting, because it is really hard to build these 1 gigawatt, 2 gigawatt. Elon wants to build 10 gigawatts and beam it up from the moon and do all sorts of stuff. That stuff is really hard, and that is where you get all the attention on are you increasing our power prices? How much water are you using? Are you compensating us through your property taxes enough?
Are all you guys in Silicon Valley getting too rich? At the smaller scale, it is a really productive use already on the land. It increases property tax for the region without any real incremental infrastructure at scale. We think that is a really interesting opportunity. I think the third one beyond the co-located, and we will bring Google, and we will bring hyperscalers, and we will bring neo clouds and invest aggressively together, is just that we really see it as an opportunity to drive new renewables. In a lot of places beyond, we will partner with a hyperscaler, and they want 500 more megawatts, and we will build 500 more megawatts. We are also seeing opportunities within our portfolio where somebody is building a data center, an SPP.
Maybe it is in the same state, maybe it is one state away, but they need a whole bunch of renewable capacity in order to make it work. Can you build some battery storage and sell us the capacity rights for that battery storage over the long term? That is a pretty interesting opportunity that probably would not have existed because the utility is probably not going to contract for that directly. That is an opportunity that probably did not exist a few years ago. In several markets like that, like in Arizona, has a similar program, right? They know that everyone who wants to build large scale data center, of which there are a lot, is not going to be able to co-locate, a gigawatt of solar and a gigawatt gas plant and whatever else they need.
They have sort of a subscription model, where you can go and find Longroad and say, "Hey, Longroad, can you build your solar and storage projects in a slightly different part of the grid? We will connect you with the data center developer." All of a sudden you can accelerate what you are doing. You can do more faster.
It is advantageous for the data center developer, advantageous for us, and it is advantageous for rate payers. We are pretty excited about these opportunities. I do feel like it is changing every day, as we said, right? By the time I finish this talk, there will be some other development item in one of our markets that I missed. We see just an incredible amount of opportunity, and it just goes to, we have been investing in these options for a long time.
In the case of Milford, we have spent 20 years there, right? We have spent 20 years there, and we have plans to build new stuff each of the next three or four years, right? That new stuff is solar and battery storage and transmission and digital infrastructure. I think it just goes to, it is a very long-term business.
You have to invest, and you have to create options and let those options grow. Here is a little highlight of what we see as the current footprint. Yellow dots are the operating fleet expansions, and the green are kind of greenfield opportunities where we have already contemplated renewables, but we have not built anything there yet. We have a lot of opportunities that we view as pretty mature. On the left-hand side of the page, you see there, in green are the grid load, the potential compute.
And then in black, we show sort of the expansion renewables that we hope will help drive that growth. I feel like I went a long time. I am going to hand it over to Pete. Give me the stage. Thank you.
All right. Thanks, everybody. What I want to do is build on what Paul and Charles talked about here, but really from a financial perspective and maybe help you connect a few dots in terms of how to value the company. It is always a question that we get, JB and Matt and Mark Flesher, so I will see if I can help you with that. Just to get into a little bit, Longroad is an integrated developer, independent power producer in renewable space with— I like to think of it as an option on digital infrastructure. What I mean by that is our base plan, we are going to double the company over the next 3 and a half years. That excludes anything that we do on digital infrastructure, right? So that would all be incremental growth.
It is a really kind of nice option on the pipeline, the portfolio, the asset base, the capabilities of the team, all of those things. Down below the holding company, we have got an OpCo, 6.9 gigawatts by year-end 2026. That is going to produce about $540 million of run rate EBITDA. I am going to talk about run rate EBITDA. It is a question we get a lot, right? Of like, what is that? So we will get into that. Jason talked about the contracted nature of the cash flows. I think something to touch on there is we talk about interest rates and how does that impact that. Isn't it kind of bad to have all those contracted cash flows there in a market where interest rates are going up? We certainly don't think so.
One of the offsets to that is just what are the residual value of these assets look like? These are 40-year-plus assets, and we are contracted for the first, on average, 20 years. So you have got a lot of recontracting available to you, and if rates are up, then power prices are up, and that is a nice offset. Now, that is a long way off. I get that.
But that is some offset to what you are seeing in the sort of the existing asset base. On the development side of the business, and maybe a good segue to that. In a higher rate environment, provided that the PPA prices move with that and we are still kind of able to make the math work, and there is still demand at that higher level. Now you have got upside on refinancing that you didn't have before.
If you create these deals in a low interest rate environment, all the sort of the upside is kind of sucked out of it on the front end. The game used to be in the super, in the ultra low rate environment, build a project at 11% levered return, but nobody wants to hold it at 11. So exit it at 9 or 8 and make that spread.
Now you've brought profit in on the front end, but it has to be a develop-to-flip model, because not interesting for an investor like Infratil to own that asset. But now all that lifts, right? Now we can generate levered returns on a build and hold in the mid-teen range, maybe a little bit higher than that. Now you've got an asset that you can hold as well, and you've also got a ton of refi upside in the future.
I think that's the silver lining to a higher interest rate environment, and we got into this business 20 years ago, and I remember because I got my first mortgage back then, 30-year mortgage, and it was 6.6%, and I thought I got a good deal. Now, world's going crazy because mortgages are back at that level. But really, it's just kind of getting things back to normal. Bottom of the page here, the investor base.
We're excited that Infratil is now our largest investor. They've been with us since the beginning. They've been incredible partners. It's been 10 years. We're looking forward to the next 10 years, Jason. The equity piece, that Infratil piece, that bottom line, that's the most important part of the capital stack, but it's actually the smallest. We do about 90% plus other people's money, OPM, we call it.
Where we're bringing in other investors into the project to finance these deals, and you've got to have access to that capital. That's just how the market works in the U.S. If you're good at raising this capital, and we are, then you're as competitive as the biggest guy, and you're certainly more competitive than the small guys. The products are on the left side of the page here, tax equity. That's a way to monetize tax credits. I'll talk a little bit more about that. Then the right side is the debt part of the capital stack. Then the middle is the opportunity to sell down assets at the project level. We've done a lot of that historically.
We'll probably do it here and there in the future, but the base plan that we've laid out has us retaining ownership of 100% of the projects for the next three years. We'll see how it goes. This is a tool, but it's a tool that we don't have in the base plan at the moment because we've got a plan that assumes we retain ownership of all of the assets.
The other thing to mention here, back to the rates and the rate environment, is what's our exposure to interest rates when you look at this 90% piece of the cap stack? Everything on the operating and construction portfolio is hedged from a rate perspective, and it's hedged over the duration of the debt. We put a piece of debt on. It's scheduled to amortize off over the life of the PPA, power purchase agreement.
You are kind of lining up my contracted revenues with my debt, and you get those things to fade away together. The rate is hedged for that entire period of time, and we hedge that at financial close, start of construction. You take that risk away for the operating and the construction fleet.
The development side, it is a little trickier. It is harder to get a rate offset in place on a development asset, but we have found a product we really like, deal contingent hedge, where once we get a PPA, in most cases, we are able to lock in our interest rate as well and take that risk away. Once we get that PPA and that revenue line is locked, then we are trying to lock in all of our input costs with cost of capital being the biggest one. That is how we manage that. All right.
Let us get into the business a little bit more and the numbers on a detailed level. Run rate EBITDA, $540 million of run rate EBITDA on our operating and construction fleet. That 6.9-gigawatt fleet. What is run rate EBITDA? Why do you guys keep talking about this? Really simply, it is the annualized EBITDA or unlevered cash flows that our operating and in-construction projects create. Very simply put. It excludes tax credits that have been allocated to tax equity. When you look at our value stream, and I use that word intentionally as opposed to our revenue stream, our value stream, we are paid for producing electron or capacity, the ability to produce an electron, like in a battery. We receive revenue, cash revenue for that, from whoever the counterparty is, our customer.
On top of that, we qualify for a federal investment tax credit or production tax credit, as the case may be, and that is an incremental stream of value. I am sure there will be some Q&A at the end about how does this business perform after that is gone, and I think those are good questions, and we are happy to talk about it. And we should. But let us just stick with the here and now, and understanding those two components. Here, I am only showing that revenue side of it. The cash flow side of it. And we will do $540 million run rate on this 6.9-gigawatt fleet. But the calendar year, this calendar year, we are going to do $290 million. It is a pretty big difference. And why is that difference?
Well, you can see the components, and you start to put this stuff into annualized revenues, and you can see how it stacks up. We got a good full year of production on our existing fleet. Some of those operating projects are only partial year this year.
They are going to be full year next year. We have got projects that are in construction. We have got projects that are going to go into construction between now and the year-end. And you got to capture the true earning value of all those assets to really start to apply multiples and make sense of the business. Here is another look at it. $540 million, again, of run rate EBITDA for our 6.9 gigawatt opco. But if you look at how that comes on over time, calendar year 2026, we are going to do 290.
It actually takes three years for all this to come online. But the value has been crystallized today. If we went to JB and said, "Here's our growth plan," he'd be like, "Get out of here. This stuff is all done. This is not the growth plan. This is growth that you've achieved." Because it's just a matter of time before these projects come online. And why is that? When we get through the development cycle, and we get to the start of construction, that's the big de-risking event. At the start of construction, the risk is kind of out of the asset.
A great evidence of that is at the start of construction, we'll put in place non-recourse construction loans that convert into term loans, where lenders are coming in for 90% of our cost at SOFR plus 150 basis points, and their only recourse is to the asset. That's it, if things go wrong. That's how well-structured these deals are from a risk perspective, that you can source that kind of capital, put all the recourse on the asset. That's why the market in the U.S. doesn't really perceive, and that's why we, as managers, don't really perceive a lot of risk in the construction cycle. It's a lot of work, and we don't take that for granted. We've got to execute. But from a risk perspective, it's a totally different thing from a development project. Okay, so that's the opco.
Let's talk about the devco side of the business. Here's a slide that Charles showed you before. He showed you in megawatts, I'm showing you in dollars. Give you a sense of the capital needs of the business, and I want to talk about how we're going to fund it. 2026, that goes to the operating and construction portfolio, so that's kind of part of that 6.9 gigawatts.
We've got a little bit of work to do between now and year-end to get all the remaining gigawatts on board. It's about 1.1 gigawatts left to go. But that stuff's kind of in the bag. That's sort of easy for me to say. I'm down here in Sydney drinking flat whites, and the team back in U.S.A. is drinking McDonald's coffee, and they're working their tails off, right, to get that stuff done.
But they're really just in the final throes of getting the project financing documents in place. So it's a lot of work, again, but it's pretty low-risk work. So now, 2027, 2028, 2029. That's about a $10.8 billion CapEx program. Excludes the data centers that we've got in front of us. Most of that is going to be sourced with third-party debt, tax equity. We've got an expansion of the existing holding company debt facility, and then there's an equity component. We need about $800 million of equity if we're going to retain ownership of all this. Some of that we have today, most of that we're going to go and get as part of this capital raise that Paul mentioned. Here's another look at kind of sources and uses, how to fund the business.
I want to talk about this one in the context of tax equity and tax credits, because I am going to try to help everyone connect the dots in terms of value. Okay, so, $10.8 billion of CapEx. Tax credits allocated to tax equity will fund about $5 billion of that, so 40%, 45%. Those tax credits are incremental to EBITDA. When you say, okay, I got tax credit over here, but I got tax equity over here, I think of tax credit as kind of the asset, the value stream.
You can think of tax equity as kind of the liability, and they offset. Now I am left with a net CapEx number of $5.8 million. How do I pay for that? I pay for that with EBITDA, with unlevered cash flows that the assets are generating, and that is also how I earn my return on equity.
If I kind of piece it all together, at the end of the decade, $1.2 billion of EBITDA and $8 billion of net debt. I take away tax credits that are allocated to tax equity. I take away both of those things, and I have got an apples-to-apples comparison. The net debt number has been adjusted to say, here is what the debt balance is going to be at the end of construction. Once the dust has settled on everything, this is what we expect the debt balance to be. I can start to do some math, and I can apply a multiple to EBITDA to get to enterprise value. I have got a net debt number that makes sense, that is apples to apples with EBITDA, and I can back that off to get to equity value.
If you looked at these numbers a year ago, and you compared them to where we are today, and you held your multiples constant, we have got an equity value that is about 50% higher than what we talked about with you here last year. It is really fun to come down here and report those kind of results and report this sort of business model and business plan that we put together.
We are going to double the size of the company in the next three years. We have got data center upside on that, and we have got an incredible team to do it with. We are really excited about the opportunity. I think one of the things that we are most excited about when we think about the sort of the vision, the mission for the company, is really to be part of the solution, right?
When you look at the U.S.
The need for electricity in the U.S. and to be able to deliver that to our customers in a competitive way, and for the greater good, right, in terms of those electrons being renewable. That's a real kind of motivator for the entire team. With that, let's see what we have for questions.
Pete, click the
Ben. Ben Crozier, Forsyth Barr. Can you just go back to that slide where you sort of talked about the amount of projects you have, say, for next year versus what you'll actually deliver? Can you just talk us through sort of what goes into selecting the projects that you go forward and why some of them fail or don't sort of reach the construction hurdle?
Yep. What was the slide that you wanted me to?
The electrons coverage slide.
Oh, the coverage.
Yeah, for the most part, it is not that they are going to go away, it is that they are going to be pushed out. That is generally the case. The revenue contract is certainly a big piece of it. The interconnection timeline is also a significant piece of it. We talked a little bit about the interconnection challenges. It is a moving target. We have in the U.S., obviously a lot of generation in the queues, a lot of generation wants to get online.
Problems you have that is an ongoing battle is several projects in one part of SPP, just pick a market. They apply. They come online. They are all 3 or 500 megawatts. You get the transmission upgrades required from the utility, and that number is too big. 1 or 2 or 3 or 4 of those projects is dropping out.
2 drop out, maybe you do not need as much transmission upgrade or maybe you do, and it is just split between 2 parties. Then how long does it take for the utility to build those upgrades? The supply chain stuff we talked about has meant that a lot of the utilities are taking a lot longer to enable the upgrades that are required to bring on the new generation. So most of it is timing, and I would say those are the 2 biggest reasons why a project might hit that kind of blue area versus one that is going to be pushed off. It is generally those 2 in our
I think a good example this year, like this current What happened in 2026 is we had a project in California that was in our plan up until maybe three or four months ago. It became very clear that we were not going to get to the finish line with a counterparty, by the end of the year, with the revenue counterparty. So we shook our portfolio around and what else is in there that we can do, and is there anything that we can bring forward? What we are bringing forward is a project in Hawaii that has been under development for a fairly long time. It has a PPA. We are just waiting for kind of one little permit that we think we can get before the end of the year.
That is the kind of bingo or the juggling that we are doing to make sure that we continue to hit the numbers.
Grant Swanepoel? Can I get at the back, Grant Swanepoel?
Can you hear me? With this capital raise of NZD 750 million going external, I am just interested on what sort of pricing you would be putting on that. When I reference that to your March independent valuator update movement to the June one, where your independent valuation went down 4%, I take it rates are rising. Since March, you guys have doubled your cadence of build-out. Does that imply that the independent valuator is seeing your new build regime as NPV zero?
Well, I think we're going to find out what the market thinks of the value when we go to market, so that'll be a great external value point. The independent valuator has their approach. It's probably a little bit different from how we think about the world. One piece of light I'll shed on that is the independent valuator heavily discounts the pipeline. So they probably, these Amargosa projects, for example, they probably value those at $50 million, just off the top of my head. Not much, right, relative to what those things are going to be worth once we get those in the ground. So I wouldn't sort of read too much into that and their approach, other than, just know that the new growth is heavily discounted under the way that the appraiser approaches things.
Thanks.
Yep.
Suraj Nebhani, Citigroup. Just one quick question for Charles on the data center plans. Can you just touch on what's likely to come up near term? I guess a related question is on the levelized cost of energy slide. I noticed the combined cycle gas turbine costs were lower than solar and battery. We are seeing a lot more of those projects going up as well. How do you think about the competitiveness there, for renewables?
Yeah. I'll start with the LCOE one. I think the point that I tried to make, right, this is the whole country, so I think that's one important point. The sort of average of the whole country for the most part, like the places that we're operating in, let's just use Arizona, right? That'll be on the very low end of the solar side. Battery storage side, we expect the cost curve to continue to come down. Then you look at the gas combined cycle side, $51, that's probably like Texas or Alabama, right? Only not places necessarily that we are competing in order for them to build that cheaply, right? Because there are higher costs associated with a lot of the places that we're operating in, right? Higher labor costs, higher transmission costs, et cetera.
Obviously, places like California, it's not going to build a lot of new gas, or Hawaii, as Paul's point. So, yeah, it's very different. The U.S. is a big place. Every state's a little different, so I think it's hard to generalize in that respect, but we feel really good about the competitiveness of the markets we're investing most heavily in. Then, sorry, your first data center point or question was? Just around what was. Yeah. Yeah, maybe just go to the last of my slides, Pete. Yeah, so, Texas has been the most aggressive place, right? The most activity. Obviously, we have one of these is a large operating campus where we feel like we have an advantage. The second one is not an operating project, but a late-stage ready-to-build, I'll call it, development project.
It has solar ready to build, it has battery ready to build, it has gas basically on site, ready to build if you wanted to build gas there. Texas, I think from a speed perspective continues to be an incredibly efficient place. Permitting requirements are effectively nil, right? They are sort of, it is permit by right. You do not have a lot of the same restrictions as you have in other places.
Obviously, the Public Utility Commission is trying to weigh in on the ridiculous amount of data centers in the queue, so there will be some external factors affecting things. Utah, I think we are excited about because, like I made the point earlier, right? We have been in Milford for 20 years. Our landowner, for the other dot there is the largest landowner in the state. A lot of sort of political advantages, and we feel like it is really well-sited.
We will see. There is a lot of dots on here. I could be surprised, right? There are some dots on here that are really large. You talk to some, the SpaceXs of the world, or some of these guys or Google, they want really, really big, right? Certain places sort of are more likely to be able to build really, really big and maybe explore some of the fully behind the meter opportunities, right? Which are less common, but growing as these things happen. I will probably be wrong no matter what I say, but I think that give you a little bit of sense of how we think about it. Yeah.
Yep. Thanks, Charles Spiliotis. Do you mind if I do one quick one for Charles, one quick one for Peter? Charles, just to follow up on the 5 gigawatts of near-term data center opportunity. Just can we dig into the potential economics a little bit more? Specifically, would you still need to buy or lease the land? What sort of CapEx per megawatt or total CapEx is it for the 5? The last one is what sort of yield on cost then are you going to get on that CapEx? Maybe I will just quickly follow on to Peter, if that is all right. Just looking at your long-term projections at $1.2 billion EBITDA versus the 14, you are kind of implying like 80,000 per kilowatt, I think, in terms of the yield.
Just thinking back to what Jase was saying earlier in terms of trying to PPAs going up and you guys trying to go more towards merchant, I am just wondering, is that conservatism or is that like that Amargosa asset mix that is bringing it down? Or because you are kind of assuming no increase in yield in future?
Yep. Do you want to go first?
Sure. In terms of the way we think about it, the scope that we would likely have, I think we're thinking NZD 3 million to NZD 10 million per megawatt is probably the portion that we would be responsible for, and that's incremental to the renewables. Hopefully driving incremental renewables, and/or battery storage as well. So 3 to 10 out of 40 to 50, which is what we sort of think about for a large scale campus today. Obviously, we're not buying chips in that example. Then, a lot of the other HVAC cooling sort of stuff that is cost intensive.
If you can give us bucks per megawatt.
Yeah. So we don't focus too much on EBITDA per megawatt as a metric. We focus really on I think because our mind is really on the asset-by-asset level, and what do we need to earn on any particular asset? They're all unique. Some have lower cost profile, higher generation, but maybe in a market where power prices are lower, some are the opposite of that.
Hawaii would be the extreme example of high power prices, but high input costs. So what we're trying to do is really put deals together that work and some might be much higher kind of dollar per megawatt, EBITDA, and some are going to be a fair amount lower. We're targeting per unit profitability at $250 a kilowatt. So that's really what we kind of drive the team toward. Historically, we targeted $200 a kilowatt.
Jason gave us a hard time about that because we kept outperforming it, and it's like you can only get away with underpromising and over-delivering for so long before-
Nine years.
That is really how we sort of manage the business. Whatever falls out in terms of NZD per megawatt of EBITDA is really just sort of a consequence, or an output. It is not really a driver.
G'day. Jason talked earlier about taking more merchant risk. Can you talk to that specifically around the appetite for that? Secondly, adjunct to that is what it means for funding structures for development?
The second part was for funding?
Yeah. You are saying you are taking merchant risk rather than a PPA-
Yeah. Right.
as your capital stack and potential amount of equity required?
Yeah, that is sort of the point, right? It is easy to say, let me take more merchant risk, capture more upside. Challenge with that, volatility in your annual numbers, which nobody likes, and the second part is it requires a lot more equity capital to undertake that strategy.
We have not seen the need to do it. We have been able to find PPAs that we think are really attractive, and therefore you can get where we are not taking a discount to market, right? We are seeing PPAs that are good market rates, and then that allows us to get a much cheaper cost of capital on the project. That is kind of the math that we do. That sort of always leads you to, let me take the PPA. I think the merchant exposure is really on the back end. That is where we have got a fair amount.
And these PPAs come off over time. This Milford project, the wind project that sort of got us started out there in Utah, that we have been able to expand so much around over the years.
That was a 20-year PPA. We actually developed that project back at First Wind, and 15 years ago. We are getting to the end of that deal. I think that is probably, practically speaking, the bigger opportunity, is the recontracting. When you get to that point in the life cycle of the project, it is de-levered. You do not have to make this choice of, I give me more merchant exposure, but I am going to have less debt, therefore more equity. You have got a de-levered project, and you have got kind of a lot more degrees of freedom in terms of how to recontract it.
Stephen Hudson.
Thanks for the presentation, guys. Just on the equity raise, would you contemplate sort of equity coming in at a project level, if that is possible, rather than the whole co-level?
No, we are trying to raise capital at the same level that Infratil and our other shareholders are investing.
No, that is clear. You mentioned just helping us in terms of what the independent valuation might look like post-2030 if you switch on and off the tax credit. Just interested in that.
Yeah. The tax credits that are in the plan through 2029, we are going to get those tax credits, right? So we will receive those. There is execution that we need to do to get those, but those will not go away. The question is really around the tax credits is this business viable once those tax credits are gone? That is the sort of the threshold issue.
We believe it is. We are investing in 2030-plus pipeline. The reason why is there are a few different things. First is, if you look at what is actually going on in the tax credit environment, the battery portion has a tax credit through 2033 start of construction, 2037 in service. So you have a ton of runway on the battery, and that was very intentional if you look at how the legislation was drafted.
When you look at our pipeline and our forward CapEx program, about half of the CapEx program is battery. So we still have a tax credit for about half of what we are doing. It takes you to the other half, right? Can you provide a product that is cost competitive on an unsubsidized basis? The projections from all the experts say, well, you cannot. We are really competing with natural gas. We get a lot of questions, nuclear and coal. Coal plants, they are just not competitive on an economic basis. It is really that simple. Nuclear is so far out, and it is much more expensive. So it is really kind of solar and battery versus gas, and that is where we see the industry going, and we compete well with gas.
It is really complementary technology in terms of enabling new renewables to penetrate the grid.
But I think the point to make, though, is in this market where the supply-demand dynamic is so strong, we are not competing with gas because customers need it all, right? If you talk to a data center developer, hyperscaler, they want gas, they want renewables, they want storage. So we are not even competing. They just want to get access to the actual They want to know that you can actually deliver the project within a certain, not in 2035. Like can you give me something in 2029? That is valuable, whether it is gas, solar, or anything. That is what is important.
Okay. I think that's a good place to end it on.
Okay.
Thank you very much, guys.
Yep. Thanks a lot.
Thanks, guys.
Kia ora and good afternoon, everybody. For those of you that haven't met me before, I'm Nick Judd, recently appointed CEO of all of two weeks. Joining me on stage I have Kieran Byrne, recently appointed CFO of all of two weeks. He has been with the company for longer. Richard Mooney, who's CEO of our EonFibre business. We actually thought we did quite a good job of the presentation last year, but it seems like we've been downgraded to the after-lunch slot. Maybe that was a bit of misguided on our behalf. The other term that I heard was we're the warm-up act for Greg Boorer and the CDC team. Either way, we're looking to outperform this year and see if we can get a better ranking slot next year. We'll see how we go.
Today, we're going to talk about a few things over the 40 minutes, then obviously leave some time for questions. There's a couple of messages that we want to reinforce. One, why we believe we're winning in market and the external data that points to that. We want to talk through, obviously, the strategy and where we're at with the change of CEO. As you can expect, there will be some tweaks that we will make, but we want to be clear that that is an evolution of strategy, not a revolution of strategy. That's because of the great work, obviously, that has been led by JP and the board over the last seven years, and how we've transitioned from Vodafone through to One NZ, and having that brand so well established in market.
We'll obviously talk a bit about RANCo, which I know has got a lot of interest and a lot of questions so far, and Kieran will touch on that. He'll also touch on where we're going with our network aspirations and why we are starting to invest in what we see as a driver of growth for us through Horizon 2 and Horizon 3, and to give you a sense about how we see that playing out. Last, and certainly not least, it's a good timing for Rich to join us on stage and talk about EonFibre, which has now been running as a separate business and how that business is shaping up.
He'll talk through, I guess, how we're seeing that market play out and some of the opportunities that we have, which are now coming into focus quickly and the opportunity with that being a separate business. Just touch briefly on the executive team, because I am fortunate that the exec team is already sorted, and that's a real credit to JP and Jodie King, our previous Chief People Officer, because we were able to take talent within the organization and fill the roles. Kieran, you'll remember, for those of you that were here last year, was our Chief Technology Officer. He stepped into the CFO role. That's enabled Sharina Nisha to step up, and Sharina's a 30-year veteran of our company. She has previously or most lately been leading our T-One transformation project and doing a fantastic job of that.
She has absolutely earned her seat at the table, and we are incredibly proud that she is joining us. Then, we have had Sonia Fernandes join us from a Chief People Officer perspective, who joins us out of Washington. She has spent time in New Zealand, Australia, and across the Asia Pacific, and brings really good experience about organizational transformation and how we face into the impacts of AI and automation, which we continue to lead on, certainly in a New Zealand sense, and how we go about it. Rich Mooney is sitting out there by himself on the side, and that is because he actually reports into, obviously, his own board, which is made up of both owner shareholders and independent directors. I am not going to spend any time on that apart from two numbers.
One we are incredibly proud of is we were awarded earlier this year, for the fifth year in a row, the best network in New Zealand by Ookla. While there is different measures and maybe some competitors have a different view, we unashamedly and very proudly believe we have got the best network, and that is proved by the performance stats that we look at and see. Our team have done an incredibly great job of building that from where it was pre-Infratil ownership to where it is today. It still takes the majority of our CapEx investment. As a result, we continue to invest in that and obviously build out 5G capability. The other number I want to talk on is one that is actually going to change tomorrow, which is our retail store number.
We have got 58 today, and as of tomorrow, we will have 59 when we launch Tauranga City again. Why that is important to us is because it is a really integral part of our sales story and our sales momentum at the moment. As we launch stores, we are outperforming every business case that we put in place. As we refresh stores, we are getting better service metrics, better sales metrics. It is probably a story that we do not talk about often, but it is actually quite unique to a lot of other industries where people are shrinking back to digital only. As handsets get more and more expensive, people want to touch and feel them, and people still want human context. You will continue to see expansion and investment in retail stores as we go forward.
The next slide, actually, I will not talk to because it sets up a bit of an agenda for the day and probably a checklist as we go through each of the areas. What I would say as a headline sort of statement is, we are very pleased with our performance today in what has been quite a challenging market. We are pleased with the investments that we are putting into areas like AI and automation, some of the value we are starting to see through that. We are pleased with how our transformation project has gone, and we are in the third year of the three-year program of that. Those investments will drive increased earnings and cash generation in future, which we obviously know is incredibly important for our owners, as Jason talked to this morning.
This will be somewhat familiar to you, because it looks very similar to our strategy on a page. As most new CEOs do, we have this chance just to stand back and reflect about how we think through our strategic priorities. For us, the purpose and the ambition haven't changed. We still want to be the best at what customers value the most. What we have tweaked is we've changed the pillars. The pillars that we had last year were best network, best solutions, best people, which were actually capabilities that we have in the organization. Where we're at now, we want to be really clear about how we win and how we drive future value growth and future cash generation. The first pillar on this is actually embedded in that AI and automation investment that we're putting in.
We want to be a much simpler and a much smarter One NZ. We know we can drive significant efficiencies at a scale and pace that we haven't been able to before because of the complexity of our business. We've got confidence that the skills we've built through the POCs and some of the value that we've created through AI initiatives are actually now going to enable us to be able to scale efficient growth and efficiency and cost over the next couple of years.
One of the aims for me in the next three years is that we drive really aggressively around radically simplifying the business. I don't think that this is an option for us, and it's not an option because I think business models are going to start changing so much faster with the way AI is now playing out on our landscape.
We have to try and throw off some of the anchors of the past of being a 35-year-old telco company and have the flexibility and agility to roll with that. We've got confidence we can do that. We will be leading with an AI-first human where it matters methodology, and some of the skills and capabilities that we've built over the last two years will really help us in that radical transformation that you'll see.
The second one, best place to buy a phone, you might think is a bit narrow. What we have learnt and what we'll talk more about through the strategy is actually we know that if we get the phone and that phone purchase, that we have customers that are stickier. They move up the product stream in terms of in the value stream from a mobile phone product perspective. They churn less.
More importantly, it actually becomes our doorway into how we win their home. Off the back of it, we can win the broadband proposition. We can win their companion plans. While it seems focused on the phone, it's actually not. It's about our unlock through the front door about how we win that home and get much higher customer value out of the home rather than the individual. We'll talk a little bit more about that today. Lastly, Kieran will touch on this in some depth. We want to be a network of choice in both Richard's EonFibre and Kieran's space. Previously in technology, we have an ability as we go forward, we think, to move from the best network to a network platform that will actually drive the next wave of revenue growth for us.
The investment that we put into that is quite important and central to how we continue to grow. While we are obviously unashamedly targeting to win a market, we know that we have a core responsibility as well as one of New Zealand's largest employers to make sure that we are focused from an environmental, social, and a governance perspective. We are very proud of the progress that we are making in this space. We have got a small team that drives some really good outcomes, and there are a couple of highlights that I just want to call out. Probably the single most important number on this is the GRESB assessment, where we moved from 79 out of 100 to 95.5 out of 100 over the last year.
What that signifies is it signifies that the time that we have spent on purchasing certified renewable energy to reduce our Scope 1 emissions, the e-waste programs that we are running, the focus we have on inclusive workplace and diversity, and the focus on partnerships with Iwi and others is actually shining through in how we are measured and tracked from an external benchmark. It is a number that we are very proud of in terms of particularly the movement we have made. In addition, and aligned with that, we also have Te Rourou , our foundation, which we probably do not do a good enough job of talking about, that has been around for decades and focuses on youth and how we can advance a better outcome for youth through digital enablement and other activities.
That program has had some really great success stories that we need to do a better job of amplifying. We are very proud of that. Before I guess, move into talking about how we are winning a market, it is probably just worthwhile touching on market dynamics at the moment. When we spoke probably 12 months ago, we had just come through the. So six months before, we had seen the trade tariffs come on.
New Zealand had started to look a little bit better, then it got the wobbles again. Same story happened earlier this year, where things were starting to look better post-summer. Then obviously with the advent of the war, New Zealand got the wobbles again. Anecdotally, certainly the feedback we are seeing through our customer base is that there is actually more positive momentum starting to shape up.
Probably will not be helped a little bit by the election going on at the moment. Actually, when you look across the board, and even through probably the worst of the sectors such as construction, what we are starting to hear is actually that there are some green shoots everywhere. We are seeing this in some of our footfall numbers. Our Apple iPhone pre-sales just went live on Sunday, and we are up double digits on last year, which is good, particularly given the handsets have gone up in price again. That is partly, I think, related to our proposition in market, but also is a good indication of market overall. We are seeing more positive signs such as, I think, Monday's announcement that net migration is starting to turn around.
While there's been really subdued growth in the overall market for the last two or three years, we actually have expectations that there is going to be a better level of industry growth over the coming years. From our perspective, we don't need that to be exponential. We don't need migration to go back to sort of '21 to '22 and the start of '23 as it was. That's not built into our assumptions around how we think we can get to a mid-tier, mid-30s EBITDA margin over time. I guess, why do we keep saying we're winning in market? For us, we obviously look at our internal financials, and we can see we're doing pretty well in comparison to our competitors.
The data that you see on the screen here is one of the better industry benchmarks, which is IDC data that we all contribute to. I think it's important to pull out a couple of points in here because I think they reinforce our story really well. That is for the last two years, we've led in overall mobile growth, and we've captured on the One NZ brand alone almost 60% of the overall market growth through that two years. That doesn't take into account the fact that in the MVNO bar, we're capturing a good chunk of that as well. When you put that on, we are far and away capturing the most significant portion of growth in the market versus any of our competitors.
That shows in the market share graphs that you see up top on consumer mobile revenue share, where obviously every year we've had a 1% increase. It's hard to move the market by a percent, so it doesn't look a lot. What it does mean is in that pay monthly segment, we are now neck and neck with Spark, when we previously had quite a sizable gap. We have the momentum in the most valuable sector of the mobile category. I'll talk about what underpins that shortly. Obviously, one of the elements that we have quite unashamedly led on has been price increases. For 10 of the last 12 quarters, we have seen ARPU growth. In fact, this was up to March, and I can tell you for June, we've also had another quarter of ARPU growth.
That now means that we're sitting at around NZD 51. That's market-leading, and yet it's not impacting our acquisition, as you can see from the other numbers. Our customer proposition now is strong enough that we can hold an ARPU premium, and we can acquire better than what our customers can. That strategy for mobile growth is built on 2 things.
It's a differentiated customer proposition, and it's better customer service than what we had. On that first point, you would have heard us talk, certainly last year and I think the year before, around One Wallet or phone dollars. Effectively, this is a loyalty scheme like Airpoints. What it does is that it means that as people purchase more products with us or engage with us, we effectively give them phone dollars, which they can then use on their next handset.
And that is becoming more and more important. For those of you that followed the iPhone launch last week in terms of foldable phone is now coming out. The starting price for that is now NZD 4,500. The top-end product price is NZD 6,500. Handsets are getting more and more expensive. The chip shortage is adding to that.
That's not going to change anytime soon. But again, we know that if we win that moment when somebody needs a handset, we get the value creation off the back of it. For us, that proposition is built on three things. One, it's built on that best network, so people can trust us when they join us. Two, we believe that our sales and our service is getting better and better, and I'll talk a little bit more about that in the second service section.
And three, this One Wallet proposition is at the moment, we're the only one with that in market. It becomes incredibly hard for others to mirror because the way we actually enabled that was through removing a whole bunch of discounts on our plans a couple of years ago. If somebody else wants to try and take us on on that, it would be a direct hit to their ARPUs and very challenging for them to do.
It provides us with a competitive advantage in that consumer space that's very hard for others to mirror, and as you can tell, is gaining momentum. It shows up in things like reduced churn. So we get a third less churn for people that are engaged in the One Wallet program. It shows up in us winning more broadband and companion plans, as I said.
So it's margin accretive in two ways, both in revenue growth, but also in less revenue dilution. It's a really important part of our consumer proposition as we go forward. Then the other side of it, which we again haven't talked so much about, but is actually the service side. If you went back a couple of years ago, on average, people had to call us once a year. Now, obviously, some people call a lot more than that, but across the customer base, it was once a year. Within a couple of years, we've moved that to two years. Actually, with the AI and automation side, we believe that we can move that significantly again. The ultimate aim for us is that no one needs to call us.
That would be ideal because if we can get ahead of some of that stuff through the use of AI and insights, actually, we might be calling people the other way around to say, "Hey, you've got a problem. You need to look at this." Now that's utopia. We're not quite there yet, but it's a real focus for us because we know that if people don't have to engage with us, there's less chance of problems. The net promoter scores go up, and again, they stick with us and stick with us for longer. So it's a big focus for us. In our retail environment, we've just launched Retail 3.0 in our Albany store. If you're in that neck of the woods, it's worthwhile going to see.
And effectively what we've done is we've created a whole lot of self-service portals, which are getting better take-up than we expected from the business case. What that means is two things. One, people don't have to wait in line to get serviced, but even more importantly, it means our salespeople can sell. Our sales numbers are up in those stores. It allows them to, particularly with the iPhone launch and other stuff, it means that we can engage with people a whole lot more quicker, which obviously drives bottom-line benefit. So it's a good example where a focus on service drives good outcomes from a sales perspective also. The last piece that I want to talk to is just enterprise, and this is one area of the business where it's challenging and remains challenging.
So, we're continuing to see quite aggressive propositions in the mobile space in terms of, and while the rate of decline in mobile ARPUs has slowed, it definitely hasn't stopped, and I don't think it will for some time based on some of the behavior and market. So, for us, that's meant that we've taken a different focus to how we try and win in that space. We've launched a product called Unlock, which effectively gives the employer the choice to slim down their corporate phone proposition. In return, we come in with an employee plan proposition based around phone dollars that offsets the fact that they're now going to have to pay for their mobile phone. So a double win all going to plan. The employer gets a reduced cost.
Obviously, we get the customer, but the customer gets an ongoing loyalty program that then funds their next handset, so there's not that employee disquiet over it. We've got a pretty good pipeline for this, and we believe that it'll be both a retention and an acquisition tool, and so the whole organization has swung behind it. There's been a few challenges working through the product proposition and how it lays out, and every organization is different. But all of the One NZ employees are now on it, and so we've worked through that, and they are now advocates for pushing it out. So that'll be a key tenet for us from a product perspective. Then we've made some changes to our structure in terms of the segmentation of how we think through that base.
And we've split our strategic customers from our mid-market base, and then we have business and SME. That's quite important because that top end of town requires a lot more bespoke solutions. Mid-market just has to be plug-and-play. We've got to get away from doing bespoke things in that space so that we spend less time on admin and drive more efficient margins through that.
Then lastly, in that space, we've taken all of the sales channels, and we've put them under Chris Fletcher, who drives really strong sales disciplines right through that function and is doing a great job of leading that team. So we're in a good shape with that. On the ICT side, we've got a couple of really good re-signs that are in the bank. Couple we've managed to build out propositions along with our cybersecurity business, Defend.
It is a continued focus for us, a continued challenging market, but one that we are not backing away from as we go forward. I will leave it there and pass over to Kieran to talk through the next pillar.
Thanks, Nick. We talked about a simpler, smarter One NZ being a new pillar of the refreshed strategy that Nick just talked about. Why that is really important is One NZ is like any typical telco, where over time, we have grown through acquisition. We have taken on different organizations, and we have not necessarily cleaned up the IT stacks that sit behind that. We find ourselves in a situation where we have a lot of complex legacy products and a lot of complex legacy processes and platforms that sit around that too. What we decided two years ago that we were finally going to solve this problem, and that is when we started to think about this T-One program, which is transforming One.
We kicked that off, and really that is a program around modernizing our IT estate, our customer-facing platforms, and moving all our customers from their current legacy platforms onto this new modern architecture. A year ago, I stood here, and I talked about we are about a year into the program, and we were just about to migrate our prepaid base.
The good news is the program has remained on track and on budget with successful cutovers. We migrated over 2 million active and pre-active connections from our prepaid base onto the new stack, shortly after the discussion last year. Since then, we have put our entire pay monthly base now in the new stack as well. In about a week's time, we have got our final major release of the program, which really sort of breaks the back of this program.
We will move our pay monthly products into the new stack and start migrating all those products over. It has been a really, really great project. It has been a challenge, but we have built up some really sort of strong project disciplines around transformation in the organization as part of that.
Other highlights is the simplification that has happened as a result of that. 98% of our mobile connections are now on in-market plans. We have gone from 130 prepaid plans down to six, and we have gone from over 700 pay monthly plans down to a few dozen now with a little tail to clean up to get that from 98% to 100%. We are making really good moves on broadband as well. That is across, that is not just consumer. That is 98% across consumer, business, and enterprise as well.
In terms of benefits that are coming through, we are starting to see the benefits being realized off the back of this project. We are seeing reduced failed orders, so less manual intervention around orders. We are seeing average handling time in the call center reduce as they get used to using these new platforms and systems.
We are also seeing massively increased IT reliability. We have seen incidents in IT almost reduced by half since we have started this program. Also the speed at which we are able to get to market now has been radically transformed from taking months to create new products to literally now creating them in hours. The next thing I wanted to talk about was AI. Somas stood up here a year ago and talked about at One NZ, we were going to make a big push on AI over the next 12 months, and we have.
We have launched around 50 different AI solutions in the business across different domains. I will call it, there is a lot of smoke and mirrors around AI adoption and enterprise at the moment, and it is difficult. We have experimented a lot, we have learned a lot along the way, and we have had some really good successes. I wanted to talk about a couple of things today in areas where we are really finding a benefit from AI in our business.
We are a telco, but increasingly we are a software business. Telco platforms are becoming very software-defined, and we spend tens and tens of millions of NZD a year on software development. One thing generative AI is very good at is software development. What we have found is that is probably the number one opportunity in our business for reducing material cash from our spend.
We have found that as other businesses have complex to deploy these tools and get people to adopt them and use them. What we have found is the best approach is standardizing a toolset that everyone can use and then driving productivity with top-down targets. We have created what we call a software workbench, which uses the best-of-breed tools we can find, which supports architecture, it supports software development, it supports business requirements, it supports testing. It gives these standardized tools to teams, and then we have literally gone through and reduced their budgets by 25% and forced the productivity through in that way. We have found that is the best way to get adoption in the software space.
We have still got a lot of work to do, and there is a lot of opportunity, and we will keep working on that software development life cycle as it is the number one way that we can reduce cash spend as part of AI at the moment.
That is the generative AI side. I wanted to talk a little bit about agentic as well, because I think 12 months ago, there was a lot of talk about agentic AI, but it just did not work, and it certainly did not work in the context of a business like ours of mission-critical infrastructure. Since about, I would say, April, May this year, the capability has been out there to really take advantage of agentic. I am going to talk about one solution here, but I really want to talk about in the context of what we are going to do next.
I want to talk about this network concierge AI agent. What that does is that takes a whole lot of telemetry off our network. You can basically give it the mobile phone of someone and a problem they are having, and it will go and diagnose that problem. That is something that an engineer would do. Typically, it would take a certain amount of time.
We have now got an AI agent that we can put that information in, and it will come back, and it is saving the people who are doing those diagnoses about 75% of the time. On its own, that is not going to transform the economics of our business. As we go forward, we are thinking about how do we start to really redesign processes around human and agent interaction and all the data and automation associated with that.
We are a process business as a telco, and so we are now going through a methodology where we go process by process through the business. We start to design out what that work looks like in an agentic world. We build the agents, we build the data products, and we redesign that to get the value out of that. We have just started that process now, and that is going to be a big focus of the next year as the complete process rewiring of our business around agentic technology. The other thing that is not on this page, which I will mention as well, is we are starting to experiment with our own in-house AI as well.
We bought a couple of GPUs, and we have got our own open-source models on them, and we are starting to actually deploy that on real use cases inside our business.
That has sort of two advantages at the moment. It not only saves us token cost from Frontier Labs, but it also, we can start to think through if we can do this to ourselves and start to put these workloads on our own environment. We have got a set of enterprise customers who are equally trying to solve the same problems that we are. By trying this on ourselves and understanding what we can do, we are going to see if there is an opportunity for us to support our own enterprise customers with their own sort of sovereign AI solutions. The next topic I wanted to talk about was around our network and how we see value being extracted from our network in the future.
Because we see a lot of change happening in the world at the moment, and we want to be the ability to make the most of that change that is occurring. There are a number of trends happening around us which give us a strong conviction that there is a growth opportunity coming that if we can pivot what we do in the right way, we will be able to capture. Really that centers around this AI revolution.
We talked about generative AI and agentic AI in that previous section. When we start to make the move into physical AI and devices and sensors and robots and drones around this world, they all need to be connected back to a physical data center. A telco is obviously in a prime position to be the bridge between that physical AI endpoint and the cloud intelligence that it needs to talk to.
It requires us to think about things differently. We need to think about different types of connectivity for different applications, which is something we haven't done before. We also need to think about how that connectivity can be consumed in different ways through third-party innovators and ecosystems rather than through the direct sales that we do now. We're starting to work towards this, and we're starting to prepare ourselves for this future by thinking ourselves more than just a network but more like a digital connectivity platform that can sort of empower an ecosystem of developers to use connectivity to deliver some of these productivity solutions. One of the foundations of that is our 5G standalone core, which we're currently in the process of deploying. With that capability, we're going to be able to do this new technology.
We're able to slice the network up into different lanes for different applications for some of these new and emerging use cases. What that actually practically looks like is you've got a traditional telco model where from device through the infrastructure, we hardwire products through there. Everyone goes through the same lane in the network. All the innovation's done in-house and we think about different sort of technology, satellite as sort of resilient complements to our network. What we want to ultimately pivot to is a connectivity platform where the capabilities of our network are exposed through APIs to an ecosystem of developers who can take them and do really good things with that connectivity and do the selling and the productization and the innovation for us. We're moving the innovation out of the inside of us through the outside.
We also want it to be infrastructure-agnostic. We want to expose connectivity products where the customer does not care whether it's satellite, whether it's wireless, whether it's a fiber. They just want a connection from A to B, and we choose the most appropriate inputs for that. I want to give a practical example of where we're actually making some progress in this domain. Halter is a New Zealand business, which a lot of you from over New Zealand will be familiar with. But for those that aren't, it's an agricultural sector business that creates collars for farm animals. Those collars do a few different things. They can obviously monitor the health of the animal. But the real interesting use case is geo-fencing.
They buzz the cow in each direction and what they can do, the farmer can control where the herd of cows goes to and effectively, without fencing a property, can move them around the property and get much more sort of yield out of the property. It's a really interesting concept and a huge productivity driver for farmers, obviously. What we've done is we've actually partnered with Halter because in a lot of these instances where it doesn't make sense to build fences in these remote farms, there's actually not mobile connectivity. So we've created them a solution where, when there's a mobile network, the cow collar talks to the mobile network. When it's not there, it talks to the satellite network. So from the collar's point of view, it's just got two different networks depending on where it is.
It delivers the same solution back to the data center. What is great about this is exactly what I alluded to on the previous page. We have not gone away and talked to farmers and created a product for a farmer. We have gone and done some work with an innovator who has bundled connectivity into a product that is delivering a productivity outcome for the customer.
They are effectively doing the selling for us. They are taking our connectivity, consuming that, and then taking out. Now we have tens of thousands of cows on our network, which is growing every month at the moment. I see this just as a thin edge of the wedge. There is going to be dozens of these opportunities where if we can find the right partners who can take connectivity and innovate off it, there is a massive growth opportunity for us.
Now this. Oh, sorry, go back one. This is not an opportunity that is going to require hundreds of millions of NZD of investment and some sort of wild idea. We have already done this.
We have built this platform already, and it is already powering our retail and wholesale business through APIs. It is a cloud-native platform built on GCP with our own development team. It gets us away from expensive global SaaS platforms. It really is world-leading. So much that another telco has actually licensed it off us, and we have got more interest as well for this platform. It is powering our wholesale business, which is doing really well. As you can see from the chart there, we have taken 60% of the net ports in the wholesale market over the past 12 months. That is strong ARPUs and margins.
As Nick Judd said, we have got a strategy to really premiumize our main brand and really focus on that premium pay-monthly consumer segment. Wholesale is a great way for us to address some of our competition with the Skinnys and the 2degrees of this world who are focusing more at that low end of the market. We are also launching new products on this platform. We have got UFB we have just launched, which is fiber consumer broadband. 5G FWA is a new product that we are launching. With our 5G stand-alone deployment, we will be able to create 5G FWA products which look much more like fiber products. We will be able to have much more assurance around speed and throughput on these plans.
That will allow us a new wave of growth where through our retail and wholesale business, we are able to go after broadband connections with a new fixed wireless, highly capable product. The final thing I wanted to talk about, which I am sure everyone is very interested in, is an announcement we made a few weeks ago about a RAN, radio access network, sharing agreement with 2degrees. Purpose of which is to basically improve the efficiency around deployment of mobile infrastructure in New Zealand. A bit of information about that, and there are certain things we will not be able to share today, just to pre-empt some of the questions, but I will try and answer everything I can. It is a separate entity, 50/50 JV between us and 2degrees, providing managed services back to each of the parties.
Really important here, but our core networks remain within the servcos, within the 2degrees and One NZ servcos. Spectrum remains within the servcos as well, as does the satellite assets, and all the backhaul as well. What that practically means is we are running different networks. We are going to share certain infrastructure, but it is non-differentiating infrastructure that sits on poles.
You will be able to take a One NZ and a 2degrees handset. You will be able to stand the same distance away from a pole, and you will receive a different experience because the smarts behind our network, the spectrum, the core network remain in our businesses. We share the things that are non-differentiating. It is going to enable us to deploy much more efficiently, deploy faster, get access to new technologies in the radio network at a faster space.
It does not affect at all the independence and the competition that we drive between One and 2degrees at the retail layer. I am sure we have done the research, but this is very consistent with international precedents. There is a lot of movement towards these infrastructure sharing agreements. It just makes sense.
The days of seeing three poles standing beside each other out doing exactly the same thing does not feel like a sensible way to deploy infrastructure. This is a move to do that. Obviously it needs to go through Commerce Commission approval in New Zealand. It has been submitted, the application at this point. We are expecting that process to take six to eight months. We are aiming for completion of this pending, obviously, approval of that process sometime in the first half of next year. That is our cover.
I am sure there will be some questions on that at the Q&A, but I will hand over to Rich at the moment to talk about EonFibre.
Thanks, Kieran. What I propose to do is just to give a quick update in terms of who EonFibre is, just to remind everyone in the room, and then talk through some of the market trends that we are seeing and then how we are performing in line with those market trends. EonFibre is one of the largest infrastructure owners in New Zealand.
We have over 11,000 kilometers of terrestrial fiber and a range of subsea assets in and out of New Zealand. We have subsea assets connecting the North and South Islands. Oops, sorry, I just realized I have gone too far. Just flick through. Yeah. Just so you got the map. We have a range of subsea assets that connect the North and South Islands. We have an ownership stake in the subsea system that connects Auckland through to Sydney.
We have capacity and IRUs on all of the subsea systems in and out of New Zealand. We specialize in providing high-capacity bandwidth infrastructure type services to a diverse and growing customer base. Everyone from One NZ to all of the local New Zealand telco operators, through to all of the hyperscalers, neo-scalers, CDN players, and international carriers.
FY 2026 was our first full year of operations. We separated from One NZ at the back end of 2024. Since then, we have migrated all of our direct customers away from One NZ to EonFibre. We have invested in standalone customer-facing systems, further cementing independence from One NZ and ring-fencing all customer data. We have entered into a long-term agreement back into One NZ for the provision of fiber services into One NZ, and we have completed an investment to upgrade our national optical network.
Effectively, we have a brand-new optical network up and down the country. What that has given us is a high-capacity network up and down the country, connecting all the major cities in New Zealand, all of the data centers across New Zealand, and all of the cable landing stations in New Zealand. What the chart on the right shows is, as well as connects up all the existing data centers, actually, our fiber network is very close or pretty much adjacent to future data center zones. These are locations across New Zealand that have been identified by Invest New Zealand as suitable locations for future data centers and AI factories based on their proximity to renewable power, grid connections, and substations allowing them to get stood up relatively quickly, relatively low seismic risk, and then also fiber connectivity.
But increasingly, as I am sure you know and have seen in other markets, the data centers are going to the power sources, not necessarily where the fiber is. We just happen to be adjacent to all of those data center zones. EonFibre is really well-positioned to support New Zealand in that next phase of digital infrastructure growth over the coming years because of our national network. In terms of trends, a lot of those trends we have discussed earlier on today, but what we are seeing in New Zealand is that AI or certainly nascent AI demand is coming through. Cloud migrations is driving significant increases in bandwidth requirements from our customers. What is interesting is where that bandwidth increase is happening. It is happening between data centers, across the national networks, and across subsea networks.
What is different is to prior growth and demand, a lot of that was access-driven at the access layer. A lot of this demand is now growing in between data centers across the national backbone network and across the subsea. These are trends that New Zealand is not at the forefront here. The U.S. and Australia are seeing the same trends or certainly accelerated trends. Locally, both Telstra and Vocus have invested or are investing significantly on their terrestrial fiber networks to cater for that demand. New Zealand is increasingly viewed as an attractive destination for future data centers due to a number of key advantages. Firstly, the abundance of renewable power. By the end of 2027, over 95% of all power generated in New Zealand will be renewable.
Secondly, its location, so it is relatively close proximity to Australia, and the fact that increasingly, the global subsea traffic is moving to this region. New Zealand is effectively on the trunk of subsea traffic that goes from Asia-Pacific through to Australia and New Zealand and onwards to the U.S. The third reason is around climate. If you have ever been there, you will know it is colder and typically wetter, both of which are conducive to data centers. Then the final point, I think, just around why New Zealand has not been at the forefront in terms of data centers up until now is because historically, when the hyperscalers have looked at New Zealand, latency was an issue in the past.
What AI has done is change that location equation because there is a wide range of AI applications that do not require low latency, and so those customers are now looking at New Zealand as a viable destination for those types of workloads. All in all, just in terms of the context of the market, how is EonFibre performing? FY 2026 EBITDA of NZD 64 million, and growing at over 10% per annum, both historically and on a forward-looking basis. We are currently in the midst of delivering a large hyperscaler subsea contract in Q4 of FY 2027, and the full-year benefit will flow through to FY 2028. In terms of the drivers of the demand, I think there is a distinction between what has driven our demand over the last two years or so versus the forward-looking.
Over the last two years, that growth that we have seen has been a mixture of One NZ and external demand. The One NZ growth has come from increased mobile backhaul as One NZ deployed more sites and upgraded more sites to 5G, which required higher capacity. Whereas the forward-looking growth that we have got in our plans is pretty much all external growth driven by the trends I have just been talking about. If I was to summarize, EonFibre is a leading bandwidth infrastructure provider with unique terrestrial and subsea assets, and is ideally positioned to capture the next wave of growth in New Zealand as AI and data center connectivity demands continue to scale. Thank you very much.
I would just like to clarify, those 10,000 cows are not captured in our market growth number, just in case you are wondering. No. I guess why are we so confident at the moment? One, just to summarize, we know we are winning a market. We know we have got a differentiated customer proposition that is landing with customers.
So we are confident in what we have got and how that is playing out. Two, we have got the strong support of our owners to invest in the platform that we believe will drive the next wave of growth, which Kieran talked through. Three, we have actually built muscle and capability through the work that we have done on AI and automation that now we believe we can drive a much simpler organization at a pace and scale that was not achievable before AI.
Rich talked to EonFibre and the opportunity and the tailwinds that we have behind that business, and we are situated really well to capture on that. We are very confident in that business plan, in fact, in exceeding that business plan. Will actually talked about this in his presentation. The disciplines that we are helping to build in the organization around capital management are getting better and better, and it is a constant focus for us. We know we can still get better at that. Every year as we go into planning, we try new methodologies, new approaches so that we are getting clearer on every dollar investing, driving a return back into the business and being crystal clear and holding people to account to that. When you weigh all that up, that gives us confidence in these medium-term targets.
It gives us confidence that we can continue to grow our cash generation, and it should give our owners confidence that hopefully that dividend check continues to grow. Apologies, Mark. Over to you.
I was going to say, I think it is the first time in a number of years we have not had a mention of the New Zealand Warriors or some sort of quiz that have required a jersey at the end.
We have not lost our passion for them, by the way. We still believe they are going to win. Just a small hiccup last weekend.
So we will happily go to questions. Ben.
Just first one on broadband. Obviously, we have had in New Zealand, all three of your main telcos have been declining in terms of total broadband subs. You do not split out fixed wireless, but can you sort of give a breakdown of what fixed wireless has been doing? Then with 5G fixed wireless, what is your sort of targets on the back of that? How much can you grow fixed wireless by either connections or some monetary value?
For a start or you want me to-
You go. I will start.
Yeah. We haven't publicly talked about FWA. It's one thing Spark has done very well, actually, when we look at it. They did a good job of getting that and holding onto that base. We had a go at it a number of years ago, and we didn't do a good job of holding onto that base. We saw enormous churn, and the customer proposition was poor. With the build of the network now and the AI smarts we've got, we're confident we can put a really good customer proposition in front of it. We've got a base there that, obviously this is not new acquisitions, it's actually migrating customers that we've got today across.
The margins that are available, I won't go into in depth, but needless to say, it's pretty obvious from some of the pass-through chargings that we get through the LFCs, which are pretty transparent. So there's an undoubted margin benefit for us if we do that. The counter to that, and being completely honest, is New Zealand probably has one of the best fiber networks in New Zealand, and it was incredibly good rollout. So it's not going to be easy. We need to be quite targeted and quite focused, and we're actually spending a bit of time now just starting to build our customer proposition and what that looks like and how we change some of the perception that we probably had. So, we've got a bit of work to do.
It needs to align with as we roll out the 5G core network and actually have the right customer experience coming off the back of it.
Yeah. I think the flattening of the fixed wireless opportunity is down to 4G really being saturated. As we're moving to a 5G world now, we've got 70% plus of the population covered with 5G. We've got this 5G standalone core technology coming and a lot of 5G spectrum. It's a real opportunity to drive the next wave of growth in there. I mentioned the presentation, but we can now create a much more fiber-like product. So actually a real competitor to the low end of Chorus's product, and perfectly sufficient for 80% of the population's needs. Now it's about creating that proposition and targeting those customers, because obviously there's a huge margin advantage for us as well.
I think the other thing just to note is the reason we haven't been able to go really hard at this already is the prices of the modems have been very high.
That's come down a lot now with global rollouts of 5G standalone. We've actually got all the sort of conditions are coming back together to have another go at this.
Thank you.
Hi, Nick. It's Phil here from UBS. Just a couple of questions from me. One, how do you measure the economics of the One Wallet? The second question I have was just, you might not be able to answer this, but over the medium term, what do you think the CapEx to sales ratio would fall to as a result of RANCo?
The second one, we will not answer in detail in terms of, but obviously it helps underpin us going to 11% or below from a general guide. The first one, we actually spend quite a bit of time on that because we look through the customer lifetime value of the product. The way that the accounting side of that works is we take the actual cost of that in an AR hit up front and then obviously get the margin accretion over the lifetime value of it. It is something that we are continuing to monitor and tweak because this is, again, down to the benefits of AI. We can now do a lot more targeted offers.
One of the things that we did not talk to was the fact that we do not necessarily have to go above the line with as aggressive offers because we can actually make more targeted offers in behind the scenes. The discounts off handsets, hopefully, over time, enables us to play less in that space, which obviously we have to fund because the margins on those handsets are so low. We look at it through the lifetime value of the customer, and effectively, we correlate the offer up the front with what the expected term of that IFP plan is and correspondingly the monthly earn that we get off it.
I think you have been let off lightly. Thank you very much, guys. Appreciate it.
Thanks, all.
Thank you.
We're going to invite the CDC team to come on stage with a short video that's going to be played first, and then David and Greg will join us.
Government, hospitals, utilities, banks, businesses, communities. The services that people rely on every day are visible. The digital infrastructure behind them rarely is. Yet working in the background, it keeps our countries connected and progressing. CDC builds and operates data centers for government, essential services, and nationally significant organizations. Put simply, we provide the critical infrastructure for critical infrastructure. That's a responsibility we don't take lightly. CDC's data centers are different. We take a long-term approach to how our facilities are planned, built, and operated. Every part of a CDC data center is engineered to operate efficiently, reliably, and securely, with accountability and sustainability at its core. Because what we build does more than support technology. At the local level, every one of our data centers creates opportunities that extend well beyond the campus walls.
First, through the people and businesses that help build it, then through the skilled teams who operate it for decades to come. From trades and engineering to operations and security, CDC supports long-term jobs, skills, and economic activity in the communities where we operate, and the impact reaches further still. The digital infrastructure we provide supports essential services, enables how we live and work, and helps our society and economy thrive. You may not always see it, but that's CDC's role, quietly enabling the outcomes we rely on every day. CDC, progress secured.
Thank you very much for coming along today. Today, you are going to hear from myself and David. David has been with us now for just past his probation period, so he is relatively safe. I think notwithstanding his poor choice in grape varieties in his wine preferences, he is not a bad bloke, and he is doing a fantastic job in that role. I want to talk about lots of things today and really keen to get to the end and answer as many questions as people have.
One of the main things I want to talk about is the differentiation of CDC relative to the rest of the industry, and that is on multiple levels, which we will talk to. I feel very privileged to be right at the heart of a lot of the conversations around digital infrastructure, AI, energy, all of those things.
I do feel for a lot of people in this room because even if you are only one step removed, 5% or 10% out of the center of the universe type of thing, it is actually really difficult with so much noise out there, to believe what is real, what is not, where it is going, how the geopolitical overlays regulation, government regulation, what is real, what is not, where the risks are, all of those things.
Rest assured, we live and breathe this all day, every day. I can certainly see, after being in the business for 20 years, I can see through the noise to see what is real and make sense of it. We continue to do a pretty good job of that along the way. When I talk about the points of differentiation, there are a few key pillars.
A lot of the folks in our industry, they might not necessarily even understand a lot of the things that I am about to say. The reason for that is people have a real short-term focus in many ways at the moment. There is a huge amount of speculation. We are not speculating. We are not working necessarily on a real estate project by project basis. We are thinking about a digital platform, and the value in that platform is longevity, is the retention of customers and the velocity of creating new customer contracts with our existing customer base and how quickly we can grow that over time. That velocity is really important. At the center of the velocity, we start with security. The world is not getting any safer. It is getting more dangerous all of the time.
Geopolitical, cybersecurity, all of those types of things, and all of the customers are craving the same sort of security protections that government has enjoyed for 20 years. I did not know at the time, but I actually did something that in hindsight was really smart, but it was really hard, is that we started at the really hardest end of government. Government is the hardest customer because they really have high standards from a security, regulations, the audits, reporting, all of those elements. The security was always the biggest hurdle. We were lucky that we started, or unlucky, but lucky in hindsight, that we started with at the intelligence agency end of the spectrum. Because if you can satisfy their requirements, you can satisfy everyone else's requirements.
That now has gone through all of civilian government, then into the critical industries of Australia under the various critical infrastructure legislation in Australia. A lot of the organizations, if they were disrupted, they would have a similar impact to the disruption of Australia. Those organizations are now flocking to CDC, they are all the biggest and best kind of organizations in banks, utilities, transport, logistics, even all the research organizations, education, and the likes. That is just another pillar of our customer base, it has all been based around security. People fail to recognize often that the government is the biggest consumer of IT products and services in Australia. A lot of the other organizations that are in our data centers continue to serve government through our ecosystem.
Our government penetration in the market is even much bigger than people first realize. A lot of that is down to the security that we have. In the future, cloud computing is so key to the delivery of a lot of the services on behalf of government and also enterprises, and is completely integrated, and they need to have the same or similar security settings as well. In the future, that proximity to the most important data in Australia is going to be super important for the AI deployments that we are having. AI, over time, that will also require the same or similar security settings that we have for all of our other customers. Security, it has been the first starting point for us, and it continues to be super important.
The good thing about security, if you try to retrofit security on a footprint-by-footprint basis or building-by-building basis, it is quite difficult, quite expensive. It is impossible, really, because of disruption. If you build it as part of your DNA from day 1, and you do it at scale, you are actually not incurring a significant or material cost overhead to build to the highest security standards, which gives you the ultimate optionality with customer choice in the future. Security is great. The availability element, people still talk about 5 nines and whatnot. We guarantee and deliver 100% uptime for our customers. As I stand here today, in 20 years, we have never had to pay out SLA penalties or anything of that nature. The availability is super important.
Number 3, optionality, I believe, would argue has been the biggest contributor to CDC's success and is a massive point of differentiation. When I started the business, I may or may not have said this to you people before, but when I started the business, I hated data center people, because data center people give you a list of constraints that you have to work towards. I am an IT person at heart, and I am just learning this data center caper. The data center world, it is completely out of step with the way technology evolves. People, they build real estate, and they think in real estate terms.
They build a piece of real estate, and then over time, that piece of real estate becomes progressively and increasingly at an accelerated rate, disconnected with the technology that it needs to support because it is built for a point in time, and we all know that technology is rapidly changing. When we started the business, we did something crazy, which is we reticulated liquid in data halls 20 years ago so that we could closely couple cooling to IT equipment.
We could also adapt and change footprints almost infinitely to changing technology over time. What we also could do is over time, we could also do much higher densities in the same or similar footprints. That really attracted the attention of the most important research organizations in Australia. Then the research organizations started deploying high-performance computing infrastructure in our facilities nearly two decades ago.
Then 15 years ago, we did our first direct liquid cooling to the chip supercomputer deployment inside a CDC facility. That was really cutting edge. The thing about supercomputing is like Formula One. What happens in Formula One is a lot of the innovations of Formula One find themselves in passenger vehicles 10 years later or maybe less. Today, if I think about it, supercomputing is still Formula One, but passenger vehicles is all of the AI deployments. We have been doing that type of deployments for 15 years. The whole world shifted, and not a lot of people have this experience. Our facilities, even the oldest one today, 18 years old, is still capable of supporting the most sophisticated computing infrastructure in the world today.
All of our facilities, all those 20-odd that are operational, and the ones that we are building will be. They are certified by Nvidia to the highest, the next generation computing infrastructure that will be deployed. In that 20-year period, almost 20 years of operation, we have had 0% churn because our underlying facilities have always been able to adapt to changing technology over time. We have all of this liquid cooling experience, where that is a really rare element in the world. As a result of that, we have worked closely with all of the leading hyperscalers to, in many ways, co-develop all of their liquid cooling architectures and deployment methodologies and commissioning activities.
CDC today is the home of one of the largest direct liquid cooling AI footprints of most of the hyperscalers, and that is all in Australia today and live, and sets us really apart and puts us in a great position for the future. That optionality, the ability to adapt to changing technology over time, do not underestimate that, because it is really, really valuable, and our customers love it.
The biggest mistake people make in the data center industry is to think that it is set and forget. This whole notion that you build a building, you sign a lease, and then that is a mature asset, fully leased, and there is nothing to do anymore. Well, that is called an office building. It is not a data center. A data center, every rack, every row, cabling, it is an ecosystem that changes every single day.
If you do not have a closely coupled relationship in terms of the success of your customer inside your facility, if you do not have the ability to adapt on a rack-by-rack basis to different technologies, different cooling architectures, air cooling to liquid cooling, different ratios within each data hall, then you will get business if people are desperate and there are no other electrons available, but you will not be first choice.
CDC is the first choice. We could sell every electron that we have as a result of all of this. The real challenge now is just getting electrons into the pipe at an accelerated rate. The ecosystem is important and again, differentiates us significantly. 96% of our revenue comes from high investment grade counterparties. There are not too many data center operators in the world that can say that.
The ecosystem, we do not want to be all things to all people. We do not want to have 500 or 600 customers. We only have about 100, 110 customers today. They are the most important customers across government, across the critical industries of enterprise, also the hyperscalers. There is a handful of neo clouds as well that we work with closely with Nvidia and others to support. Again, 96% of our revenue comes from that ecosystem. It is really a funny thing.
Once you get an ecosystem up and running and you get all of these sites connected with fiber and you have the same standard operating procedures, the same security domains, all of those things across all of those facilities, then the more government and enterprise you have inside your facilities, then it is the big flywheel, then the more attractive you are to more and more cloud and AI computing deployments.
Then the more AI and cloud computing deployments you have, then the more valuable you are to the government and enterprise customers. Even all of those, there is a slide later on, but all of those different target market segments of the market, they are all growing. Some are faster than others. There is always this notion that at some point in time, cloud was going to cannibalize this other business, it is just not the case.
You heard it here tonight with One NZ. Today with One NZ, the fact that they have deployed their own GPUs. There is going to be a renaissance in hybrid computing, because there is going to be a trade-off between the creation of your own tokens locally with your own models, combined with leveraging on a case-by-case basis, the larger models that the people need for different types of workloads.
There will be lots and lots of different permutations of that. The nice thing with CDC, we are not just one data center with one customer or a data center platform with five or six customers or a data center with 500 customers. We are actually perfectly positioned to win on whichever way technology goes into the future because the most important consumers and the most valuable companies and government agencies are already inside the ecosystem. The ecosystem is super important.
We will talk about this later, but the social license to operate, as well as a grid connection and all of those things, that is becoming the most important thing, the most difficult hurdle, and community engagement. Governments and councils and things, they are hands-off. It is all care and no responsibility. They have left it up to us to lead the charge. On that, CDC has impeccable history, track record of doing the right thing for a long period of time. It is well documented that I started the business in 2007. At that time, we were at the end of the millennium drought, and I did not think that it was the right thing to do to build. There were no regulations. Nobody was asking for it.
I just did not think it was the right thing to do to create a new industrial consumer of drinking water at scale when we are all showering with buckets so that we could water the garden. Nobody asked for it back then, but it has become really important now. We are the only data center of the scale that we operate at in the world today that does not use any water, and that is a key differentiator. When you start to talk about, with local communities, and I have stood in front of communities, and I have stood in front of councils, and they have got the pitchforks ready to do me in.
You actually start talking about some of these elements, and then the fear goes away, and by the end of it, they are asking, "Well, we have actually got more land if you wanted to build more data centers." The sustainability is really important. Since 2007, because government was our customer and, government, that particular year, 2007, Kevin '07, was when the election was fought, and won on the greatest moral challenge of our time, which was climate change. We immediately decided that we would offer renewable energy because government has got to eat its own dog food. Government was consuming renewable energy. We have been offering renewable energy, zero water solutions with really low energy efficiency ratios for 20 years.
The new rules and regulations that government is trying to bring in, some proponents in our industry and new entrants that are coming in and may or may not be here in five years' time, their businesses might be here, but they may or may not be here. A lot of these people, they do not really have this long-term view of how the world works. As a result of that, a lot of people are trying, the data center industry are trying to talk the standards in government off a cliff to the lowest common denominator. I am really supportive of making the bar high because, one, we have already operated at that level.
But two, we cannot lose the trust of the communities. They have really valid questions to ask, and we should be able to answer those questions, not dismiss them as a hurdle to doing business, because they have valid concerns, and we need to think about what our facilities are going to look like and operate in 20 years' time.
That is where we are really, really careful with site selection. In a perverse, self-punishing type of way, it is actually a nice thing in some instances because it is another barrier to entry and actually helps us remove a lot of the speculators from the market, which is the sustainability element. As of today, the only data center business that I know of that is 100% net carbon zero in 2026 across all Scope 1, 2, and 3 emissions, which is pretty cool.
That gets you a long way when you are talking to the regulators around what good looks like now and into the future. The scale is important. Our scale today, when you think about some of the facts and figures that we will talk about soon, the scale gives us the ability to circumvent, in many ways, a lot of the challenges of organizations that are thinking about their businesses on a project-by-project basis. That is access to long lead time equipment, making commitments earlier, all of those things because we have a production line of long lead time equipment that we have ordered yonks ago, so that is here on time. Then we are just pushing out, but thanks to standardization, that infrastructure to the appropriate project at the appropriate time. So scale has given us a big advantage.
I would hate to be starting my business. I started in 2007, then we had the GFC in 2008, which was a lot of fun. Today, I do not think you could necessarily do what I did in 2007, in 2027, just because the game has changed, the scale is different, the barriers to entry are incredibly high and relevance is important. You are not relevant unless you are talking about big scale, and you tick all the other boxes, which we have managed to do. Finally, densification. I love this because I am actually a data center operator. I am not a financial engineer. I love what I do, and I get excited about every data center tour that I do because I get to see one of my children again.
These data centers, they are just like the most massive big boy train set because they are just all these combination of cool elements that are stuck together. It is the melting pot of all these different engineering disciplines that I love. If you get all of that right, then you have a significant advantage because we are not just building a piece of real estate for a point in time. We are building something that can adapt to changing technology, customers, security, densities and whatnot. This densification, the current average rack density that we are deploying for AI deployments, for example, is 132 kilowatts. When I started CDC, it was 0.5 of a kilowatt was the average in Australia. Think about that. 0.5 of a kilowatt of power in every rack. Then I did something crazy.
I said, "Well, the minimum we'll offer is 5 kilowatts a rack," because no one else could do it because of floor density, power, and cooling constraints. Then suddenly people realized that it was cheaper to rent one rack under one piece of roof, supporting 5 kilowatts than renting 10 racks at 0.5 of a kilowatt each. We sort of got going, and others didn't have the technical levers, therefore the commercial levers to compete with us. That densification story continues. The average cloud density now is about 13 to 15 kilowatts a rack. The AI footprint, as I said, 132 kilowatts a rack, and we're deploying hundreds of these things at the moment. But next year, from April next year, it'll be 220 to 250 kilowatts a rack.
The magic here is if you've built a portfolio of data centers and it's in here, there's some numbers there. Just say round numbers because I'm not very smart. 1 gigawatt of capacity. I estimate that we could squeeze 100 to 150 additional megawatts of capacity out of that existing set of shells with a very, very small growth CapEx component because of the flexibility and fungibility of the architecture and the engineering inside our facilities. That would be a combination of simply densifying the footprint and then upscaling over the course of the contract people's capacity requirements. But it's also, Jensen's talked about moving the operating temperature of silicon from 26, 28 degrees that it is at the moment, to 45 degrees.
Now, the moment that we sort of get closer to that, and we're moving closer to that all the time, then suddenly all of this energy, these precious electrons in the world, we can redirect that because of the flexibility of the architecture that we've already got in our existing facilities from the mechanical cooling electrons into leasable revenue-generating electrons in the same physical shells.
Now, that's incredible. But I don't think people are talking about that because they probably don't really get it, because they haven't lived and breathed the customer requirements based with technology at the coal face. Certainly, there wouldn't be too many CEOs floating around that have cleaned the floors of their data centers at some point in time. That's me. The densification is a real sort of upside, and there's nothing in what David says includes any yield optimization for densification.
It's just tick, tick, and just think about how undervalued we are. If you wrap all of that up, then you have an incredible picture in front of you, which we're really proud of. This is 20 years of my life's work in front of you today. Because we've been really disciplined with the way that we've used money over time. I say to people, a lot of people in data center world think this is a get-rich-quick scheme. It's not. It's a get-rich-slow scheme. I'm just lucky that I've been doing it for 20 years. Because we've been reinvesting constantly with the support of our wonderful shareholders, and we continue to reinvest a lot of free cash flow.
There's only three data center operators in the world that have an investment-grade credit rating, and we're one of them, and probably the closest to the center of the universe when it comes to the AI revolution. What does that mean? It says those available debt facilities, NZD 11.4 billion, which we've also worked really, really hard to establish. We've got access to bigger, larger pools of capital than other people, and our cost of capital is ridiculous relative to the market, something in the order of 6%. That's great because it just makes it, as every year goes by and as every billion dollars is invested, we just get further and further ahead, in terms of our ability to execute, our ability to grow the pipeline, et cetera, et cetera, without tapping our long-suffering shareholders for additional equity.
It's going really, really well from a capital side of things. Really, when you think about it, the constraints are really simple. Capital, chips, and electrons. Capital you can eventually solve for if you pay enough risk premium, but we don't have to pay much of that. Chipsets are relatively well sorted, but electrons is where the field of play, the battle is at the moment, and we do a pretty good job of that.
Average weighted lease expiry is still 28.7 years. People think about what's the re-leasing risk? There is no re-leasing risk because our existing customers, we can adapt to changing technology. It's so much more difficult to move than to stay, and if you love the operator and they've loved you in return, then it's unlikely that people will want to move. A 0% churn rate over 20 years indicates that.
That's the old world, not the new world. The old world was storage and processing of data. That's cloud, that's enterprise, that's government. That's not the AI world. The AI world is the generation of intelligence. There is not enough supply for all of the demand for the intelligence that needs to be generated now and into the foreseeable future. Those electrons, they'll be the most valuable thing. Those electrons will continue to be in demand regardless of who plugs the computers or who's financed the computers or owns the computers, they will continue to be in demand. Don't worry about re-leasing risk. That should give you a lot of comfort, those types of numbers there. We talked about energy, we talked about water. We don't use any. Then our pipeline. We've got about 355 megawatts of revenue-generating capacity today.
There was a slide there about in, I think Jason's presentation around signing a 555-megawatt contract. I was a bit lazy. It took me about 18 years to sign 200 megawatts of capacity, and then it's taken me 12 months to sign another 800. I was sort of starting to find my feet. Training wheels are off, and we're away. No, we've signed now, it's more than a gigawatt of capacity. Again, if I had five gigawatts of capacity today, I would sell five gigawatts of capacity today, and that's what we're working on. We're very conservative here, because we've got very, very conservative institutional investors behind us. We only put into this slide here what absolutely we own the land, we've got commitments on power, and we're working through the consenting process in the appropriate locations.
But rest assured, there is a conga line of additional capacity beyond what is here. We are building 1.35 or that sort of building and operating. We have about 2.6 up our sleeve that we are trying to accelerate the delivery of. That 2.6, I probably have genuine demand, and it changes every night, because I get text messages every night about more capacity.
Probably 8-10 gigawatts of capacity that I could deploy, I could contract if we had it. That is what we need to be able to sort of work through into the future. That is in Australia and New Zealand. It is true, New Zealand, although it is more expensive to develop there because of the geological risks and what you have, you have got to build a building, and it bloody rains all the time, and the earth is soft.
That it is actually becoming a more and more attractive location, and that is one of the reasons I will talk to Contact in a moment, but that is one of the reasons why we are a bit more up and about and a bit more positive on New Zealand than we have ever been before.
Because this business has pivoted completely from me competing for business to me allocating capacity to the appropriate customers as that capacity matures and becomes available, which is remarkable when you think about it. So that is where we are operating today. It has been a wonderful journey and it is very, very exciting and there is tons more to talk about, but in the near term. So there is lots of green there. Some people might argue that the capital expenditure should be red, not green.
But I think about that as an investment, and we are getting terrific returns on every dollar that goes into the ground. But everything is on the up. People talk about it. I do not really think about it because I am thinking five years ahead. But apparently, it is interesting that we have had an upgrade in our EBITDA. So it used to be AUD 680 million - AUD 720 million , and now it is probably more likely somewhere close to the NZD 750 million.
So that is exciting, but it does not deter from what I do every day, which is just trying to keep the machine moving faster, which is great. Now, when we think about just some examples of the locations, and these buildings, these are beautiful. A lot of people think that you can have a data center that is just quite functional. I believe in form and function.
So I think you can have quite a functional thing, but it can also look great. Our buildings, again, we are building nine data centers at the moment, and that will continue to increase. The size and scale of the data centers will continue to increase. Melbourne in Brooklyn, you got Brooklyn 2 next to it. The Brooklyn 1 is the small one on the left, and the bigger one on the right is Brooklyn.
That is about 120 megawatts of data center capacity that is all now operational in recent times. Eastern Creek 5 and 6, this is a good story about densification. If you think about the buildings on the left, closest to the old Coles depot there, that is 100 megawatts of data center initially designed, but 150 megawatts of data center deployed in the same shell as an example of densification.
The IRRs on that building are off the charts as a result. Marsden Park, that is Marsden Park 1. That is our very first data center, so that we have got a little footprint there for people to connect all the comms to and all of that. Behind that, there is 550-odd megawatt data centers getting built behind it right now. There will be 720 megawatts of data centers.
I am a few days away from selling the last building, and then it will all be sold. That will all be delivered in the next couple of years, which is great. Hume 6, we are still building in Canberra because there is lots of demand. That is only probably a 30-megawatt building in Hume 6, but that is a new one that has just gone live. Beard 1, that is a new campus, and that is brand new.
Auckland Campus 2, this is the biggest data center in New Zealand, about 35 megawatts, just up the road from our original location in Hobsonville. That is just about, if not completely sold today, and that is why we have got to move and build the next data center at Silverdale, which is already consented and approved as well. New Zealand continues to go well for us, but we are looking at doing bigger and bigger developments over there as well. Full on into construction, Laverton, that is Laverton 1 and 2. It was only, it seems like the other day, that we returned soil on Laverton 1 and 2. They are 75-megawatt data centers each. We just got planning approval for the third one, which is lucky because I had already sold it.
That is three, and then we bought the land across the road where you can put another three of those. We are full noise on that. This is exciting, Perth. Perth is going to be great. Perth is going to be a big data center market at some point in the future. It has just got oodles of renewable energy. The sun shines a lot, and then it gets hot there. In the evening, you get the Fremantle Doctor comes in, and wind power is really, really good. You can get a high percentage across the year of genuine renewable energy 24 hours a day with less requirements for batteries and storage and other things. We are excited about that. It has been an interesting community engagement activity, but we have won over the community, and they love us.
We have had all the councils and things over to the East Coast to actually look at our facilities. When you can talk about acoustics, you can talk about renewable energy, you can talk about not using water, you can talk about the community impacts of what we have done in our community over the last 20 years in terms of benefit sharing, community work, charity work, sporting organizations, then people get pretty excited about us being in their jurisdictions, which is great. There you go.
There is Marsden Park. There are the data centers on the previous slide that have come online this year, and these are the big locations that we are building at the moment. That is 150-megawatt data center on the left next to Marsden Park 1, Marsden Park 2, and there will be more MPs there.
All of the stuff that we are talking about, the 1.3 gigawatts of development that we are doing at the moment, all of that is on land that is already consented. All the power is there. The only risk is execution on what we are doing. Much the same way that Longroad talked about their pipeline, we are the same. These days, speed is about everything. We have probably got close to NZD 1 billion of land, and we are working on our pipeline constantly.
There is an allocation each and every month, each and every year, sorry, for additional pipeline. As you get bigger, obviously, your pipeline wants to be bigger to maintain the same velocity and trajectory of growth. That is what we are pushing hard for. This is the new Kemps Creek site that we have got initially, I think about 500 megawatts planned there, and that is going through the consenting process at the moment.
The only real risk here, and we will talk about the contracted capacity that we have done, but the real risk here is around just the execution. On the execution, all of the locations that we have talked about, we have got the power locked in, and all of the sites that were not even in the presentation today. We are working on power two, three, four years in advance, grid-connected power, because behind the meter gas and whatnot, it might be a thing in the U.S., but it is not a thing here. Just ideologically, I think it would be difficult to get off the ground. The grid is key, and we are ahead of the curve. Because of the size and scale of CDC, we can do that.
The rules and regulations around what you have to have in place to apply for and get a power consent today, there is quite a high bar, which is another barrier to entry, of course. Long lead time equipment, because the standardization of our model, as I mentioned, we have long lead time equipment locked in. Because we have been partners with these organizations for 20 years, we give standing commitments.
We are not ordering on a project-by-project basis or ad hoc. It is more of a strategic relationship with the Vertiv and the Schneider Electric of the world. We have queue prioritization, pricing benefits, all of those things because we are underwriting, in many ways, their factories ahead of time. So we are not seeing the challenges that other people in the industry are having. This strategic engagement, we always make the hardest, longest decision is the best decision.
Those hard decisions are going into communities before you lodge your development application. The last thing you want them to do is to see that there is a data center getting built, then you are always behind the eight ball. So we are engaging with the community early, state governments, councils, doing tours of our data centers consistently.
I reckon I am doing two days a week of education sessions with people, but it is absolutely the highest and best use of my time at the moment because it is a massive differentiator. We try to find local subcontractors, where we have built incredible businesses have grown up around CDC with incredible loyalty across all the different subcontracting disciplines to CDC because we have changed people's lives and they have shown loyalty to us, and we continue to show loyalty to them. As a result of that, we go much faster than most people.
With regards to the joint venture, I think you will see more joint ventures in the future because just the velocity. It is just the velocity if you just wanted to do. It is like a hyperscaler doesn't build all their own data centers because they can't do it fast enough.
If you can bring the Contact, a great group of people leading renewable and infrastructure folks in New Zealand, and they have got a great site down at Stratford, which is a disused power plant with all of the electrical infrastructure in place. It is not like it is a greenfields location. There is existing infrastructure that we can tap into. They have got existing relationships. They have been there forever in those locations. Very, very similar to the Longroad story, leveraging existing infrastructure so you don't have those costs involved.
If that means that we can move faster, and we can share the upside and share the downside risk, and if it makes sense for us, noting that we are only going to do a joint venture if what we bring to the table is adequately valued in terms of the customers' experiences, the technical capability, the construction expertise, all of those things, and obviously the capital.
If all of that is valued satisfactorily and it makes sense, then we will look at JVs because, again, it is just about the velocity of moving as fast as we can in a world today where, again, you can sell every electron attached to a data center that you can get. If you are the right company and you have got the right track record and you have got the right set of counterparties, you can definitely do it.
With that, I might hand over to David to talk money.
Good afternoon, all. It is great to be able to come and talk to you. One of the hardest things in my job, and there is many hard things, but one of the hardest things is I have to speak after this guy. Really, what chance have I got? I will do my absolute best for you today. It is great to see some familiar faces here also from my previous slide. Thank you for coming. Starting with our contracted capacity at CDC. Our capacity is now at 1.1 gigawatt. That is contracted capacity. You will recall last time we spoke, it was 1 gigawatt. We have actually contracted another 70 megawatts in new contracts over the last couple of months, which is very exciting. Indeed, since May, just over 625 megawatts of new deals for CDC.
As you will recall, on the 5th of May, a 555-megawatt contract we announced, a beautiful symmetry of numbers. I love numbers, and it is easy to remember. It was with a high-end investment grade-rated hyperscaler, 10-year deal with two 10-year options, which we think really underscored the value that we bring to the market and the esteem that we are held in by our customers that we will be able to contract on that basis with such an impressive customer.
On the right-hand side of the slide, you can see how our megawatts turn into revenue earning over the next three years through until FY 2029. Starting on the left at 350 megawatts is what we are physically billing today, that is revenue earning today. You will recall last time that we spoke a few months ago, that number was 220.
We have converted another 130 megawatts into revenue earning over that period of time. Looking through to FY 2027, we have another 100 megawatts or so that will become revenue earning by FY 2027. If you roll the clock forward to FY 2029, by March of 2029, we will have completed and rolled out the 555 megawatt contract. Indeed, all of the 1.1 gigawatt contracted workbook will be revenue earning by March of 2029. So, a really busy and exciting couple of years that we have in front of us. What does that mean for earnings for our business? Firstly, we are talking today about an upgrade in our FY 2027 EBITDA from previously NZD 680 million-NZD 720 million, up to a new range of NZD 710 million-NZD 750 million.
There are a couple of things that are driving that. Firstly, the 70 megawatts of new contracts that we have announced, which will roll out over FY 2027 and FY 2028. Secondly, some OpEx savings in our business. Thirdly, we provide managed services to some of our customers, which in simple terms is installing equipment on their behalf and managing that process. We have been able to earn some additional margins through procurement savings, which have flowed through to the bottom line for FY 2027. So NZD 710 million-NZD 750 million is our updated guidance range. Looking through to FY 2028, we still expect EBITDA to exceed NZD 1 billion.
Very importantly, by the time that the 1.1 gigawatt is fully revenue earning and fully handed over to our customers, which is March of 2029, on an annualized basis, we will have EBITDA of NZD 2.2 billion. So specifically what that means is for FY 2030, that will be the first full year at NZD 2.2 billion, but on an annualized basis, it will be in place from March of 2029. In terms of capital expenditure, our guidance is unchanged at NZD 3.8 billion-NZD 4.2 billion, excluding land. We did make a comment earlier on about land and the investment that we are making to help build out our future pipeline. We have given some guidance or assistance on CapEx and what it costs for us to construct.
That is unchanged from when we last spoke, at mid-teen level NZD million per ICT or per billable megawatt in our businesses. As you would expect, we will come to CapEx in a little bit more detail in a moment, we are very careful with how we spend capital and where we spend it. We point it at the earliest revenue earning opportunity for our business. Because critically for us, the more that you can condense the time from when you start spending CapEx to when you hit your first RFS or Ready for Service date, the more you can compress that, the less funding you need, and the higher your level of returns. That is absolutely our focus as a business.
Looking at some of the history of CDC over the last few years, back to FY 2023, I think what is really impressive as you look at this and you think about this business over the years and the incredible growth that has happened, particularly over the last 12 months, as Greg mentioned, 800 odd megawatts. But through all of that growth and through the growth in the contracted workbook, and the EBITDA for our business and looking forward, the fundamental competitive advantages we have around our weighted average lease expiry and also the quality of customers that we have, the investment-grade customers has been maintained. All the way back in FY 2023, we had 90% investment-grade customers.
In terms of our revenue base, as we sit here today, it is still 90% investment grade, and our weighted average lease expiry is still at a very impressive level of 28.7 years. Our customer base, we started with government, expanded into national critical infrastructure, expanded into hyperscale. We have variety of customers, we have high credit strength, and we have quality long-term customers that underpin our business.
I will talk a little bit more about credit rating a little bit later on. But one of the key things that Moody's look at when they rate us is the quality of our customers, the length of our expiry or weighted average lease expiry, and the quality of our shareholders, which we have talked about today, all critical to underpin our credit rating. Funding. This has been a busy space. There is no doubt about it.
For some reason, I seem to be really popular with the banking community, and I just cannot quite work out why that is. I assume it is my sense of humor and my overall personality. But I think what it is we are actually really busy on funding at present. We have a really strong and well-capitalized balance sheet. As I sit here today, we have over NZD 5 billion of available liquidity for our group, mostly in the form of undrawn debt, plus some cash on our balance sheet. You saw on the earlier slide, we have total facilities of just over NZD 11 billion, NZD 11.4 billion to be precise.
Over the last six months, we have been very active in the markets, firstly in the bank debt market earlier in the year, and then from April 21, which is when we made our credit rating public at Baa2. I should add, we had held that rating on a private basis for the previous four years. The business is very well acclimatized and accustomed to operating within the requirements of the rating. Indeed, our shareholders have been very supportive of the rating over all of those years, including providing equity support as recently as February of this year at NZD 500 million and then back in late December of 2024. Very important to underpin the rating of our business. The public credit rating allowed us to access public debt capital markets.
We issued our first public hybrid in June of 2026 at NZD 1 billion, our first senior bond in July of NZD 700 million, and we have been active again for the third time into the USPP market just last month, which is due to close in another month or so. As we go forward and we think about funding our business, we have got guidance in the market for CapEx at NZD 4.8 billion-NZD 5.2 billion.
You can assume over the next couple of years, whilst we do not have public guidance, we have got significant delivery in front of us to deliver the 1.1 gigawatt of contracted capacity. It is important that we continue to diversify our funding base, lengthen the tenor of our debt, and of course, optimize the cost of that debt. You will see us continue to be active in the funding markets in the coming months and years.
Capital deployment, I have touched on this already. CapEx, it is really important, certainly from my perspective, certainly from our shareholders' perspective, and indeed the debt investors into our business also, that we are careful and we are considered in how we spend our CapEx. What I can say to you and what you can see on the slide is that for the current financial year, over 90% of our spend is on delivering contracted capacity with our customers.
What we are not doing is borrowing funds from the market to invest in speculative build. The majority of it is in contracted capacity. Then to a lesser extent, of course, we are investing in future capacity for our business. As Greg touched on, the pipeline is very significant for us. That is it from me. Thank you. Look forward to some challenging questions. I will hand over to Greg.
It seems like we are running out of time, but I am not sure what that means. Just really quickly, because a lot of people have asked questions around, basically right now, there is obviously a lot of interest in the data center industry. There has been tons of speculation in the data center industry. As a result of that, there has been a lot of nervousness across the community.
Government understands completely the imperative from a future economic prosperity perspective and also from a sovereignty perspective and the national interest overall, and the fact that if it was not for data centers right now, that we would be in recession. Government is completely aware of that, and they want to win back the narrative. The way they are winning back the narrative is to bring down some pretty high standards and expectations on the data center industry, which we are very, very supportive.
One of the challenges you have is that a lot of the actual rules, planning rules and everything is a state thing, then there is a federal, and it is difficult for federal to and the conflicts of federal managing state, et cetera. Everyone is having a kick and a punch at it at the moment. Rest assured, we are across all of those different conversations and have regular audiences with those.
None of this scares or worries or concerns CDC because we are operating and have operated to the standards that they are contemplating, noting that a lot of this has still got a way to go. We are ahead of the curve and in a very good position relative and very supportive of having high standards that we all have to strive up to and not having a ceiling, not a floor.
That again talks to great uncertainty, and this does reinforce the strategic advantage that we have across lots and lots of technical, environmental, commercial, borrowing. There is a lot of differentiation here. Even our shareholders are even another point of differentiation in terms of being large, open-ended investors. That is another advantage.
There is lots of differentiation. We are not nervous about that at all or in any way. We have talked about a lot of these elements. We probably have not mentioned the skills element, but we are actually leaning into with our own academy, creating our own skills to ensure that we do not have a skills shortage. We are looking ahead to what is the size and scales and skills requirements of our team in the future, and we are meeting that internally rather than trying to chase other people's employees. That is working really well, and that also develops an incredible amount of loyalty.
This goes back again, if you take a long-term view, it is an infrastructure company. Contracts are long. Sales cycles are long, all of those things. If you take a long view on most things, you will do well. Again, just sort of amplifying the sustainability highlights. There is no other data center operator that has all of these key metrics. Even the PUE, that is the power utilization effectiveness ratio.
That is on par with most data centers that are spewing millions of liters of clean drinking water into the air each year. That is an annualized number across all of our data centers. Some of our data centers are very early on, and so they are very poor efficiency at the moment until you actually build up the volume of implemented computing infrastructure inside them. That number will fall over time as well.
The takeaway there is we can actually achieve the same energy effectiveness or energy utilization as organizations using water but not use water, which is pretty remarkable. Customer demand. It is probably where people have lots and lots of questions, but do not worry about the notion of how full data centers are. There is no real capacity anywhere.
Most new data centers are actually completely contracted before they even get out of the ground. That is our experience as well. There is definitely, you can see from the chart here, you can see the volume of capital that is going into data centers. What you can see here is why, and non-AI compute, and so that is classical compute, government, enterprise, cloud computing, that continues to grow. AI training continues to grow, and inferencing is where the big movement is.
It was about three months ago, people always ask me about the AI bubble. About three months ago, for the first time in human history, the amount of compute on the face of the Earth that was dedicated to training, actually inferencing was larger. That was the first time. You will actually see inferencing getting bigger and bigger and bigger.
That just goes to the consumption of AI in economies around the world. Even though we are in early days of that, you are going to see a huge amount of growth, in inferencing over time, which just speaks to more demand. There is basically more demand across every client of CDC's, and we are in a wonderful position as long as we keep building capacity. This talks about the size and scale of what people believe, Mandala and McKinsey, the size of the market is.
It is currently the operating capacity and revenue generation in Australia is about 1.5 gigawatts, but they think it is going to be about 5 gigawatts. As I mentioned, I think whatever capacity we could bring into Australia, you could just about sell at the moment. There are a few people who have asked me, the Australian Financial Review has reported about Anthropic signing a 2-gigawatt lease with Zetta, a Singapore-based developer of data centers in Central Queensland, today apparently. In saying, what does that mean for us? It does not mean anything for us.
It means what we need to do is bring on more capacity because those types of customers have almost an infinite demand. Whether it is OpenAI, whether it is one of Elon's companies, whether it is Anthropic or indeed all of the big organizations that we work with that service that market and others, of course.
The demand is off the charts. Just today, one of the most important customers in the world is just desperate to chase another gigawatt of capacity and wants to put it in Australia. Again, that is not a demand issue. Supply at the moment is the challenge. I have talked about all the different market segments. All of our market segments are growing. This is about the growth in the last couple of years, and you can see huge volume of growth across all of our different addressable markets. As we sit here today, government is still probably about 25% of our business and growing. Everything else is growing as well, which is a really wonderful place to be. With that, we would like to invite people to ask questions. Make sure they are all directed towards David.
Now the fun part. Grant.
Greg, you talk about all this demand, and it looks like you are moving towards 400 megawatts of annualized capacity growth sometime in the near future. What is stopping you from pushing that number even harder? Is it capital constraint? Is it building constraints?
It is basically constraints. The constraints, the number one constraint, is grid connectivity. The components that you need to connect to the grid can take transformers and the like can take two years to arrive. That is two years, and you cannot do much about that. Then those components need to be compatible with the utilities standards and all those things. It is not like you can get something that is cheap and cheerful and drop it in there.
To get to that point, to order those transformers, you have got to do grid studies and work with the AEMOs of the world and the utility and get permission to actually apply that volume of capacity to the grid because we have a national energy market here. That is the biggest choke point. Absolutely, we are opening up more and more of those conversations.
The wonderful thing is you will see is that there is actually, and contrary to what you read often, there is actually lots and lots of spare capacity in the grid in Australia. The problem is the capacity is not in contested, congested markets. The capacity is in locations further away from capital cities. You will see more and more of that semi-regional and genuinely regional footprints.
We have multiple gigawatt sites under due diligence right now and working through that process. They might not be in here, but that is what I am really focused on, and that is what I think about when I go to sleep every night is how can we get more capacity faster. The grid is the natural constraint. I do agree with the Longroad guys that the fastest way to bring in new capacity is a combination of renewables and storage solutions.
Even the storage solutions have come a long way. We are even contemplating in some of these big scale locations to maybe even have a 2-hour grid scale battery backup BESS systems to complement our backup on-site generation, because that will give us runtime without using diesel generation, and also be able for us to flex a little bit at different times to help the grid in times of peak demand.
That is the constraint. In parallel to that, naturally, we are involved early now in all of the community engagement activities. So they are the sort of activities. It is certainly not capital. We are actually blessed. You saw the numbers there. We are blessed with that, plus the support of our shareholders, which is terrific. But I would want to go much, much faster. In a perfect world, I would love to bring 1 gigawatt of capacity per annum online.
That is what I would like to do. You need a certain number of electricians. You need a certain number of even tower cranes. Tower cranes are a constraint. Within all of those constraints, I think that number, 400 megawatts, I would love that to be a base, not a ceiling.
Thanks. I just want a quick second question, just on the relative economics between New Zealand and Australia from a build cost and a power cost perspective.
Yeah, it is 30% more to build in New Zealand. Obviously, the depth and skill base is a little bit lower than Australia in terms of these types of construction activities and large-scale construction activities. The closest thing to building a data center is building a hospital, and you can see not many hospitals get built well. That is the cost side of things. That is where we work towards, and it takes a little bit longer as well, just because of the additional construction activities and ground conditions and all those things.
The power contracts?
Power contracts are, I would say, comparable, like in terms of a per kilowatt basis, but the penetration of renewables is higher. With hydro and thermal, you have a larger percentage of non-carbon intensive energy sources in New Zealand than Australia. Certainly since the troubles in the Middle East, New Zealand inquiries are up.
The only reason why you would go into New Zealand is because you can get capacity there, even though it is more expensive.
No. Well, the thing is, what people need to understand, and I laugh at some of the experts writing in newspapers, because they are so off the mark it is not funny. The value proposition from an AI, the commercial realities of generating tokens versus providing electrons for people to rent data centers is completely different. For every NZD 1 billion I build a data center for, it used to be NZD 1 billion was the most.
If I had a NZD 1 billion data center, you would have NZD 1 billion worth of IT equipment in that data center. That is air-cooled computing infrastructure. But these days, with GPU technology, TPU technology, those things, then it is more like NZD 9 billion to my NZD 1 billion. But that NZD 9 billion generates a much faster and bigger economic return. The most important thing is the time.
If I can guarantee a time, then actually the input cost, I could put my data center lease costs up by 20%, and it would have a negligible impact into the cost of delivering a service to the end customer. We are in a great position, and people will pay more for certainty because even if you miss the date by a month or two months, then that is a material impact when you have got billions and billions of dollars of GPUs depreciating and not generating returns. The returns are immense if you can get the GPUs and if you have got the technical capability to deliver that as a service outcome. No, you cannot have another one.
That is it. Thank you.
Oh, you are just going to say thank you.
Yeah.
Oh, sorry.
Hi, it is Elise from Pendal. Thanks for your time. I have got a two-part question, but the first one, talking about densification, can you explain to us how that then flows through into increased returns for the data center? Then going to the second part, which is basically what you just discussed. As the return profile for your customers is increasingly getting higher, and once the GPUs are depreciated, potentially significantly higher, is there a way for you as a data center operator participate in the upside?
If you think about, we have talked about just in round numbers, if a gigawatt of data center costs you NZD 15 million to build. If you can add another 100 kilowatts of capacity per megawatt, then that additional upside, which is revenue generating, is going to cost you maybe a fraction of that NZD 15 million because a vast majority, 60% or so, is into the shell and the in-ground services. The fit-out costs are probably another 15%-20%. It is really a little bit of additional Generation, generators, chillers, a few more pumps, and then you are away. Because we already upsized pipework and electrical reticulation to be able to take advantage of that densification in the initial NZD 15 million per megawatt. What you are getting is really almost additional revenue for free. That is the densification portion.
Can you do that both through the life of the contract?
Yes, absolutely. That is the thing, they talk to us all the time, and we are monitoring, and when they get close to thresholds in terms of hitting the ceiling of the capacity that they have leased, they are really open because that is the cheapest, best way to grow as opposed to having to get another data center. They much prefer that, and it really helps them, over the course of that initial contract. That is why there is generally no quibbling or challenges when it comes to the end of their initial term and taking options and all of those things.
Absolutely on the other side, we are looking at a range of models from no risk to more risk, where if we want to take on perhaps a lower credit-rated counterparty that has got a higher profitability potential, then we might not want to risk our investment-grade credit rating to have them in our data center unless we actually participate in the profit upside because getting access to the electrons is almost impossible. If we are good enough to provide that to them, then perhaps in return, we should earn more than just the lease rate. We are looking at all that stuff.
Thanks. Could I just dig into just returns versus cost of capital more with David? Just on the returns, some of your DC peers are saying they are targeting a 10% yield on costs on hyperscalers. Just wondering if you guys have a comparable target or number that you have achieved. Then just on the cost of capital side, 6% WACC is really good. What sort of incremental cost of debt have you been getting on your recent deals? I am sorry. I am going to annoy you, Greg. Can I follow up on the contracts? You said-
Why don't you let him answer?
Yeah. Certainly. Great question. So two questions. In terms of levels of returns, I can say some things without directly answering your question. Maybe if I referred to an earlier slide that the Morrison team put up where they put a range of levered IRRs of 12%-20%. So in terms of levered IRRs , we are probably targeting at the upper end of that range.
When we look at deals, obviously every deal is customer specific. Probably would prefer not to put a yield on cost number out there, but I know that competitor you are referring to and the number they put. But maybe if I quote levered IRRs , that is where we like to get to. In terms about our cost of debt, and we showed on the slide before what we have been up to over the last six months or so.
If you were to look at this financial year, full year 2027 through till March, so these 12 months, our forecast weighted average cost of debt net of hedging is about 6%. So that is where we sit for the business and some of the earlier questions, valid concerns about rising bond yields and what impact will that have as we grow our funding portfolio in line with earnings.
We have a very active hedging program. So we hedged forward our floating interest rate cost quite a significant period of time. For this year, we are about 95% hedged. Next year is about 80%. If you went through to 2029, which is when we will have delivered the 1.1 gigawatt, we are over 60%. So we have insulated ourself, I think, pretty well from risk around rising bond yields. Yeah, so that is what I would say.
Sorry. I might have misheard you, but did you say you were about to sell the last building in Marsden Park? Is that like roughly there's six buildings, so would that be about 100 megawatts? I guess the follow-on question is, on top of that 100 megawatts, physically, what could you sell on top of that 100 in the next 12 months?
Well, it depends on what we can bring on. I would be forward selling sites as we get consenting. We might be able to sell another 500, 600 megawatts of capacity. It just depends on timing of consenting. Again, if I can consent and have line of sight, and all of the construction planning in place, if we could do a gigawatt of capacity, we'll sell a gigawatt of capacity. Right now, that last building is anywhere between, depending on the configuration of the end customer, it could be 135 to 150 megawatts. No worries.
Greg, can you just talk to your medium term?
Where is it? Oh, over there.
Over here, mate. Can you talk to your medium-term ambitions outside Australia and New Zealand?
Yeah, I think, obviously, there is tons to do in Australia and New Zealand, and building and developing your own home market is much easier, safer, more predictable, profitable, all those things potentially. But I do think at some point, if we are going to realize our potential, we are going to have to have a larger degree of geographic diversity, and we are going to have to do that because my crystal ball is not as good as yours.
I do not know if it depends, do you get a change of government and then there is momentum for some sort of data center moratorium or whatnot. That is not going to happen anytime soon in Australia today. But imagine if it did, that would be a disaster. There are lots and lots of opportunities outside of Australia where we have been looking at them for the last year or so.
We are getting closer and closer to perhaps doing something. But again, we are not going to risk the mothership in order to do something on a very much an ad hoc or opportunistic basis. It has to really make sense, and it has to be absolutely demanded by one of our great customers. Then if that was the case and we had the right local partnerships and relationships, then yep, we would do it. We would absolutely do it, as long as the shareholders agree, of course.
Thanks for the presentations, guys. Just a quick one on the AFR article. Is the 1,400 megawatts sort of it? Based on what you are saying, it is basically stage 1 of a number of stages of demand from them?
No. The particular customer you are talking about, they have given letters of intent to lots and lots of players with gigawatts of data centers because they know that not all of those data centers are going to come to fruition. Even the AFR article, the one there, it still does not have regulatory planning approval. There is still a bit of water to go under that bridge.
Queensland and federal governments are locked in a bit of a battle around how data centers are powered. There is potential legislation impacting on the government-owned coal-fired energy generation assets in Queensland. Again, there is so much noise out there. That sounds like a great article, but I have read other articles where other people have already signed leases when they actually have not signed leases. I know what is real. What we have talked about is real.
That probably could be real, but there are still tons and tons of conversations and gigawatts of demand beyond that from not only that particular customer, but from the broader ecosystem of opportunity.
Appreciate it. Thanks, Rick. Just one quick one. Just with regards to the conflict in the Middle East, obviously a sensitive subject. Can you give us some idea of where Australia's global market share could move to as a result of that?
Well, even before that happened, it was already black and white that since 2024, without question, the second-largest destination for global data center investment was Australia, after the U.S. The U.S. will continue to be massive, and Australia is already the second-largest destination for data center opportunities. India has become more significant as a result because of proximity to the Middle East, but not being in the Middle East, but also the incentives and tax regime settings that the Indian government have put into place.
CDC is about the same size as the Indian data center market today. You can think about the growth opportunities that there could be in places like that. New Zealand will get their unfair share of global workloads as well, I believe, because they are considered super favorable by all of the largest technology companies in the world.
Again, that cost to build over there, that flows through. That is a rounding error when it comes to the actual cost of the stack. I do think Australia, as long as we do the right thing and we maintain the trust of the community and the populations by doing the right thing, we should be the second-largest data center market in the world. Geopolitically, it would mean that we are in such a strong position geopolitically if that was the case. From a sovereignty perspective, economic security, national security, cyber resilience, all of those things. We actually need this compute in this country because having compute in your country now is equally or more important, it is impossible, but equally important as having your own energy generation in your country that you control.
Yep. Infratil, earlier in the day, put up the slide around the next step along was GPU ownership. Have you given any thought to owning some of that on behalf of your customers or customers ask you to do that?
We look at that regularly, I would not say no to that. That would be a bit of a departure from our core business. I think there is ways of participating in that layer of opportunity without actually being the full-blown kind of owner of that sort of activity.
Okay. I know we are just over time. Give me one last burning question before Meryl.
Hi. Good afternoon. Thank you for your time. Just a quick question on the leasing side, perhaps. You talk about passing through increasing costs of potentially 20% is kind of a rounding error. What are some of the changes you are seeing to key terms of leasing with this supply-demand challenge? Just curious to see if you are seeing any new dynamics on that side.
Yeah. It is actually really interesting. One of the most educational things you can do is go through all of the different debt investors' offices and answer all their questions as well, because they are always worried about not the blue sky opportunity, but what is going to stop them from getting their money back. I absolutely understand that because of the demand and a lot of new entrants, a lot of the new entrants are getting very much more onerous penalties in their contracts and also very tight sunset clauses if they miss dates and things like that in terms of being able to terminate contracts and whatnot.
The good news here is that because of all these customers have been our customers for such a long time, because we have operated on sort of agreed templates, order forms for such a long period of time, the terms and conditions are not really changing, and it is just really about price and inclusions and all of those things which are really easy to work through. We are being really sensible, where our lease rates are probably on the incline, but we are not gouging because I am not trying to optimize every single project because I know the opportunity in the next project and the next project and the next project strung together is, and that velocity is far more valuable than a one-time deal like a lot of people are chasing.
There is not really any impact on us, but we are also not getting the pressure to drop our prices or anything like that, which is awesome given the volume is bigger.
Okay. I am going to hand it to Jase to thank the guys and to wrap up.