Goodman Group (GMG) Earnings Call Transcript & Summary

Good day, and thank you for standing by. Welcome to the Goodman Group FY '26 Full Year Results Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. [Operator Instructions] I would now like to hand the conference over to your speaker today, Mr. Greg Goodman, CEO of Goodman Group.

Yes. Thank you very much. Good morning, everybody. Five years ago, we made a deliberate decision to position Goodman as a major global provider of digital infrastructure, supporting the rapid growth of technology globally. This meant repositioning our portfolio and operations towards large infrastructure scale industrial property and data centers, assets in urban infill locations close to consumers where demand is most durable and the assets are hard to replicate. We've executed this strategy by selling and repositioning more than $8 billion of assets over the past 5 years. We disposed of properties no longer aligned with our strategy and recycled the proceeds directly into our development pipeline. And we used our internal expertise to deliver complex infrastructure, intensifying selected sites and securing the power and planning approvals needed for data centers. Today, we have a portfolio approaching $90 billion and work in progress of almost $20 billion. We have concentrated our industrial portfolio and development pipeline in high-quality modern assets capable of supporting advanced automation and robotics. Our data center program is gaining momentum, and we are now hitting our stride. Projects are on track and customer commitments are progressing alongside the construction program. We continue to build into strong demand in supply-constrained low-latency metro markets, supporting cloud and AI inference deployments. Our data center work in progress has a completion value of over $15 billion and contains almost 500 megawatts of capacity. The global workbook comprises 10 developments across eight metropolitan data center markets, reflecting the scale of our business. Supporting all this is a specialist data center team with local expertise in each market and global capability across Goodman Group. To date, our team has navigated significant economic volatility, supply chain disruption, securing contractors and critical equipment needed to meet energization and delivery schedules. Very importantly, we have the capital in place to fund the build-out of these projects with approximately 90% held through our data center partnerships. Our recent announcement of a signed lease for the first 50-megawatt phase of a 1-gigawatt Tsukuba Tech Central project shows how we are delivering on this plan. We acquired the site in 2022 alongside a long-term strategic investment partner and secured power and fiber in 2024. In response to strong demand, we started construction for first building in 2025 with a leading local contractor. We've now secured a 20-year lease with a global hyperscaler customer. Fully fitted and operated by Goodman, the 50-megawatt facility will be ready for service in early 2028. With the project underway, this provided our customer a shorter time to market and greater certainty around that important delivery timing. Securing this global hyperscaler customer unlocks Tsukuba Tech Central as a premier data center hub in Tokyo. The Tokyo lease is one of several opportunities progressing across the Goodman power bank. Slide 15 of the presentation sets out the delivery timing and leasing status of our work in progress in more detail. With deliveries running from '27 through to 2030, we're progressing customer discussions in parallel with construction to optimize commercial outcomes. A number of projects are in negotiations with several now advanced, and we're engaging with customers across the balance of the work in progress. We're also in negotiations and engagement with customers on other sites across the power bank that are not yet in work in progress. We'd expect further leases to be signed over the coming period, certainly over the calendar year. Our capital management strategy remains disciplined. The group has maintained a strong financial position with low leverage and significant liquidity. Our gearing sits at 6.5% with $6.4 billion of cash and undrawn lines. While this gives us the capacity to progress our development program, we continue to work with long-term capital partners to provide investment opportunities, manage our risk and return. In the world where we live, capital is becoming more selective. Our investment management and capital market programs is proving to be a key competitive advantage. Over the past 5 years, we've raised more than $60 billion of debt and third-party equity across the group and our investment partnerships. Our investment, operating and capital management strategy is continuing to deliver strong outcomes for our partners. And today, we announced an operating profit of $2.675 billion for FY '26. This represents 10.1% growth in operating earnings per security. I'll now hand over to Nick to make some comments.

Thank you, Greg. I'll begin on Slide 20. So we'll first cover the items that relate to our cash-back measure of earnings, the operating profit. And as usual, this excludes the unrealized fair market value movements on the properties, mark-to-market of the hedges and the accounting fair value estimate relating to our employee long-term incentive plan. The general strength of the Australian dollar over the year had an adverse effect on the translation of our foreign-denominated income, but that was offset by gains we got from our hedging. That gives rise to a $37 million benefit in our interest line. And we'll talk more about this as we go through the numbers. Investment earnings increased by 7% or $44 million over the year. There was a $16 million adverse FX translation impact. So this was a $60 million increase on a constant currency basis. Like-for-like income growth contributed $20 million of this increase. The movements in our investment positions accounted for the remaining difference. During FY '25, we had a substantial increase in direct property holdings. Over the course of this year, however, a significant volume of assets were sold to partnerships. We contributed our share of equity alongside our partners. In addition, the partnerships added outside debt that was used to acquire the properties. So even though we had a significantly lower closing balance on our direct holdings, we owned about $1 billion of additional direct property in FY '26 versus FY '25 on a weighted average basis. As a result, our direct NPI was up by $44 million overall. The bulk of our investment income comes through our co-investments in the partnerships, and this was fairly stable. Despite our increase in investment by period end, we had nearly $380 million less allocated on a cash-weighted average basis. Offsetting this was the underlying income growth. There's scope for a significant portion of our directly owned assets to create new partnering opportunities over time. This will reduce our direct investment and NPI, but increase our co-investments in partnerships and management income. At the same time, it will provide cash to fund our expansion. Over time, we do want to grow the investment part of the business as we continue to expand the portfolio of assets under management and our share of it. Continued equity investments for development and acquisitions funded jointly through the creation of new partnerships and growth of existing ones should support this. The portfolio remains under-rented, and we're invested in properties that should exhibit further market rental growth to support the increase in our investment income going forward. Management revenue was down $147 million overall. This includes the $9 million FX translation effect. The main reason is that the performance and transactional revenues contributed $206 million this year compared to $372 million last year. The performance of the investment partnerships was higher in FY '26 than FY '25, but there was a reduced number of them eligible for calculation. Excluding the transactional and performance-related income, revenue from management services was up $28 million on a constant currency basis. Total fee revenue as a percentage of stabilized third-party AUM was 1% for the year. Our total portfolio stood at $89 billion at June. Of this, $75.4 billion was in external assets under management. Within that, the stabilized portion averaged $68.7 billion this year, and that's up from $66 billion last year. In terms of the outlook for this segment, we expect our third-party stabilized AUM to grow over time. The main driver of this in the next few years is likely to be the stabilization of the data center properties we're developing. We expect to continue to invest in warehouse properties, too. Partly offsetting this in the near term will be the ongoing refinement of the portfolio and the current self-imposed limitations on development of this type. We remain comfortable with our long-term guidance of fee revenue averaging 0.9% of third-party stabilized AUM. Our realized development earnings were up by $454 million this year. That was net of a $15 million FX translation effect. Included in the results are $734 million of operating profits related to the reversal of prior period valuation gains on properties that have now been sold. As in previous periods, we don't reflect these gains in operating profit until the transaction is complete. So those profits aren't double counted over time, we notionally offset them against the current period valuation results when we do our reconciliations. Both the volume and the mix of activities have driven the significant increase in income. Activity levels have increased materially this year. Our current WIP represents an annualized production rate of over $7.5 billion. That's up from $6 billion at the same time last year. Over the past couple of years, this sort of growth in WIP is what we've been planning for. The data center development program has progressed according to our expectations. We've also made the decision to include the full MEP fit-out on all but one of the buildings in response to the nature of the demand we're seeing. The growth in DC work has materially altered the mix of our WIP. Given the time in WIP, we require and expect a higher margin to compensate. We're also originating a significant volume of work on the group's balance sheet or in specific development partnering arrangements. That means a higher realization rate. In other words, a greater portion of the development income will be reflected in our cash-based operating results rather than a share of revaluation gains. We're enthusiastic about the prospects for development overall. Customer investment demand and our ability to service it bodes well for future revenue as well as growth in AUM. Based on the current timing of the FY '27 activities, we expect the earnings to be largely skewed to the second half. The increase in our operating expenses has been moderate -- we had a $75 million increase in net interest income. This included the $37 million benefit from the hedges I mentioned earlier, but there's also been a $32 million increase in interest earned due to higher cash balances. On average, our directly owned development assets have increased, so capitalized interest is also up by $20 million. Directly owned development assets increased significantly over the last 2 years, but that occurred mainly in the second half of FY '25. Since then, the allocation is progressively declining as we've begun to joint venture many of the properties. As a result, the rate of capitalized interest has been declining sequentially for each of the last [ 3 half years ]. Our average cost of borrowings on our loans is currently around 4.6%. But taking into account interest rate and currency hedges, the net WACD is around 1%. In the near term, the interest line in our income statement will be mainly driven by the amount of cash we have -- we invest and FX rates. As far as the non-operating items are concerned, we had nearly $1 billion of unrealized valuation gains. That represents the group's share of the $3.1 billion across the entire portfolio. From that, we deduct the realized valuation gains and deferred tax liabilities to get to the $158 million net result you see in the table. Cap rates have declined from 5.1% to 5% and market rents have increased by 0.6% overall, and that was 1.3% if we exclude the effect of Mainland China. Another customary area of difference between operating and statutory profit is the unrealized fair value movement on the hedges. The rally in the Australian dollar was the main driver of that gain. As usual, we exclude the LTIP accounting cost, but include the tested units in the denominator when calculating our operating EPS. The increase in the accounting cost this year was influenced by the movement in the security price on the ASX and the higher number of securities remaining unvested. The rise in the outstanding awards was in part the result of the migration to the 10-year LTIPs, which means that a lower-than-usual portion of the outstanding grants became eligible for vesting. A few remarks now regarding the balance sheet on Slide 21. As a result of the creation of new partnerships for our directly owned stabilized properties, our investments decreased by $1.2 billion over the year. Our share of the stabilized assets in the partnerships on the other hand, was up by $0.7 billion over the year. There was $0.6 billion of new investment of equity by the group and $0.8 billion of revaluation gains. Partly offsetting this was the $0.3 billion impact of disposals from the partnerships and $0.5 billion FX translation effect. Commensurate with increased development activity, our development holdings are up by $1.9 billion overall since June 2025. Our share of the portion held in partnerships was up by $1.7 billion as we took up our share of the equity for the acquisitions and CapEx of the sites we're developing alongside our partners. The direct working capital allocation to the group's inventory and investment property under development increased by $0.2 billion. Despite the transfer of some of our sites into partnerships, we've continued to invest into their development and acquire new ones. The progression of this part of our balance sheet is in line with our expectations to this point. We have substantially -- a substantial remaining development working capital capacity following the raising last February. When it's appropriate, we also expect to partner more of the assets we have on our balance sheet, which will give us further capacity to fund more activity as we move through our power bank and industrial developments. Our cash position increased marginally during the year. We completed three global bond issues and repaid some maturing bonds and tendered for some of the outstanding ones. We invested $2.4 billion into our partnerships, and this was largely funded out of our retained earnings and proceeds from the bond issues. Overall, we generated $2.7 billion of cash-backed earnings this year. Over $1.9 billion of this is reported through the operating cash flow statement. In FY '26, the operating cash flow associated with inventories was very similar to the operating profit from developments for this portion. This is unusual for a growing business like ours, and the difference has been significant in recent years. It reflects the sale of inventories into partnerships, but with new investments being undertaken on investment properties either directly or in partnerships. Those investments are reflected in the investing cash flow. As usual, the statutory statement of operating cash flow does not include the profits we make from the transactions involving investment properties. Some of the gains from the sales from within the partnerships have not yet been distributed, which gives rise to differences between OPAT and operating cash flow. The partnerships retain income for reinvestment purposes. This is in line with our capital management and distribution preferences. We view this as a voluntary reinvestment insofar as that we could distribute but have collectively chosen not to. The combined effect of the treatment of these gains and the distribution policy was in the order of $0.6 billion. This was by far and away the largest driver of the difference between OPAT and operating cash flow. The remaining difference relates to the timing of receipts of performance fees. We've accrued income for fees that are shortly due and payable. This is required because those revenues are virtually certain. You can see from Slide 22, we have significant financial capacity to help manage market risk and capitalize on suitable opportunities that may arise. The group and partnerships are in a strong position. Across the entire platform, we completed $11.4 billion of debt initiatives and $19 billion of derivative hedge transactions during the year. We have substantial funding capacity, and we're very well hedged against interest rate and FX volatility. And that's all from me. Thanks, Greg.

Thank you, Nick. Now in closing, looking ahead, our strategy is clear, and we believe the opportunity over the next 5 years is very significant. Large-scale logistics opportunities are emerging in several markets as customers look to consolidate and to automate. Our industrial portfolio and development pipeline provide large-scale modern properties needed to support power-intensive operations. And we continue to actively acquire and progress large-scale sites capable of providing the next generation of infrastructure. In data centers, continued growth in cloud and the shift in AI workloads from training to inference are driving significant demand in our metropolitan markets. We are building into this demand and our sites, team and access to capital position us well to capture these opportunities. We will also remain disciplined in regard to capital management, keeping leverage low, deploying capital selectively and partnering importantly with long-term capital to progress the development program. And in closing, we enter FY '27 with an attractive and substantial development workbook. We have significant opportunities across our global markets, and we're in a very strong capital position to support this growth. So for FY '27, we're targeting EPS growth of 9% on FY '26. Thank you. Nick and I will now take questions.

[Operator Instructions] And our first question comes from Simon Chan with Morgan Stanley.

A few questions. The first one is just on Slide 14. I appreciate the additional details you've given us there. But I noticed just some sites you're in advanced negotiations, some sites you're in active negotiations. I mean you would have chosen those words for a reason. Like which one is more likely? Which one is more actually advanced? Can you give us some color on that, please?

Yes. Chan, if you go to Slide 15, you see we've done it. We've made it easy. We're trying to make it easy for people. And we respect that the information around this is like super important, and we get that. That's why we went out a couple of days with Japan, Tokyo, like a major, major milestone for that project, and there's planning on buildings 2 and 3 as well there. But look, if you look down the Slide 15, you look at LAX, well advanced. Hong Kong, well advanced. And Amsterdam, we've got -- we're advancing as well. So they're the three you should focus on. But then we've got Paris, Frankfurt, Sydney, Madrid and Paris 2, which we have colored differently, and that's to shade where we are with our activities. So I think LAX, HKG10 and AMS01 focus on that. But also we've mentioned in my speech, but also, I think, in the text that there's other activities outside this we're dealing with in France, for example, on some pretty big deployments. We're dealing also in Sydney, which is not in our power bank secured yet, but also on some potential deployments. So there'll be other things around this that will advance as well, but there will probably be some of the bigger deployments on some of the bigger sites. So I think Slide 15 will give you a pretty good view of where we're going. And look, we know everyone is watching this. We know it's a very, very serious endeavor. But to put it in context, if you're not building it and you're not closing out the risk at the back end, you should not be leasing it, right? Because then you'll end up with a -- you might have to wait a little bit longer to get your deal done. But if you move too quickly, you can end up with a very, very big problem at the back end. And we are super conscious on doing this right. We understand when you're dealing with some of the biggest customers in the world, it's really important we get our track record on the right side of this. So that's why Tokyo is so important. That has been in negotiation for 6 to 7 months, probably longer actually, maybe 9. And we didn't pull the trigger on that too early until we were very comfortable on that delivery in '28.

This is an elementary question. But those years on top of Slide 15, are they calendar years or are they fiscal years?

Yes, they're running to calendar , I think. Nick?

Yes.

Okay. Cool. I just got a question in relation to FY '26 earnings. Guys, now that it's all said and done, right? It's all said and done. Can you give us a few pointers as to how to work out how much data center-related earnings was actually in the FY '26 EBITDA? I'm particularly interested in the European side of things. So any color on how to think of that would be good.

Look, I think we're not going to get specific about different projects, but it's a number of things, and Nick can take you through it. But where we're creating partnerships clearly is important for us. But that's P&L, but it's -- capital is the big driver around those activities. We've got two or three. We've got follow-up raisings with at the moment, and we'll have a couple of partnership creations during this year as well. But that is a very, very much a capital-driven activity because the amount of work in progress that you see on the page today, closing on $20 billion, that will go higher in the next year. We're going to make sure we're well, well-funded. And that's a big activity for Nick, the team and myself to make sure we're well ahead of this. Otherwise, it's not going to get behind when these things -- these are so capital intensive. But Nick?

Simon, it won't come as a surprise to you, but you know that from years gone by, there's a variety of different ways that we contract and therefore, has different profit implications, and we manage the entire book of opportunity accordingly. So I think giving you any sort of more specific detail around it probably doesn't actually give you too much information. I think what we encourage people to do is look at the correlation between activity levels and opportunity, look at the inherent profitability of our development book, look at the propensity to earn 90 basis points on our third-party stabilized AUM. Most of the other factors are pretty well under control. You can -- we're giving you the programmatic sort of timetable for the existing WIP and the completions and when income can be generated, subject, of course, to leasing. But look, I think at the moment, in terms of kind of the standing investments, out of the $90 billion is about just under $6 billion associated with data center income. Obviously, the development are a high portion in data centers. And that's going to have a correlation with the amount of income we've earned out of data center development. But giving you anything more than that, frankly, could just be misleading.

That's fair enough. And I just got one more. Greg, in your response to my question before, and I think you also printed in the preso, you talked about data center programs expected to progress throughout '27 and will increase WIP, et cetera. Can you just give me a little bit more color on that one, please? Is that suggesting that there will be additional data centers potentially kicking off or going into WIP outside of the stuff that's on Slide 14 and 15?

100%, yes, that's what it means.

Our next question comes from Howard Penny with Citi.

Just one of the questions that's a big debate. Congratulations on the lease at Tsukuba earlier this week. But one of the questions we're getting is how do the different sources of revenue from a contract like that flow into FY '27, '28 and '29. And so could you just explain to us just thinking about management fees, development returns and eventually rental income, just the sort of timing of those earnings from a typical contract like that?

So let's not characterize it around Tokyo necessarily, but Nick will give you a bit of a view on a typical what we're going to do with the data center once we have it leased and how we're going to move it into holdcos and things like that.

Yes, exactly. So I mean that is -- that property and other properties are being developed on the basis that -- or the history of that partnership, in particular, is being developed to sell. And so at the right time, we'll enter some form of contract for sale. And as you know, from past experience, there's a number of different ways and different types of ways that we sell development properties ranging from presale right through to sale on completion. And so depending on the nature of that contract, we'll determine how revenue is recognized on the development portion, and that is the development profit on sale plus any development performance fees. The development management revenue itself, that is emerged as we develop, and that's ad valorem. And then in terms of the -- once it's gone into stabilized third-party AUM, that's when our sort of 90 points type fee arrangement kicks in. And obviously, our share of the equity brings up our share of the investment income, but that will happen at stabilization. So that's typically the way we've done things in the past, and I see that and others that we're working on being no different.

And just another key debate that comes up in the market is just if you track operating cash flow over the last, call it, 10 years versus underlying operating earnings, since moving into data center development, you've seen cash flow -- more cash outflow as you've been investing into all the groundworks and substations of these data centers. And we've seen your cash flow come under pressure. But could you just explain how that has impacted both that cash flow relative to that underlying earnings over the last 3 or 4 years?

Yes, that's why I spent about 3 minutes of my prepared remarks talking about the difference between operating cash flow and operating profit. And I do so every half year and have done for the last many years. But you're right, what -- typically, the difference is exactly what you're saying. You've got a growing business. So clearly, as you're growing, you're investing -- your outbound investment, which is recorded in the operating cash flow for the inventory component, is higher than the stabilized like-for-like run rate. So if you had a business that the ins and the outs were constant and there was no other change, clearly, operating cash flow and operating profit would line up. But we've had increasing investment. And therefore, typically, it's been -- it's weighed on the operating cash flow because that's how the accounting works. And then a lot of what we do in terms of the gains that we make are reflected in partnerships and the profits that are generated in equity accounted investments. And so they're not reflected in the operating cash flow if we're reinvesting those profits. So that -- those two things are the most significant drivers of the difference over time. But ultimately, it just means that we're investing in a growing business. And that's why we buffered our working capital and equity capital in February last year. The rate at which we're -- our expenditures are progressing very much in line with what we had anticipated, so we've been planning for it. And so we're in a pretty strong capital position. And that's why our payout ratio is what it is as well, right? So we intend to reinvest long term into our assets that we're developing and continue to hold them for the rental income and potential capital growth over the long term. But we want to be funded sustainably, and that's why we retain a significant portion of our operating profit -- underlying operating profit.

Congrats on the execution.

Our next question comes from Cody Shield with UBS.

Just firstly, on the Aussie DC partnership slipping. Can you provide any detail on what's prolonging in that process? Is that on the capital side or the power side?

No, nothing to do with it. We're building it. I don't know if you've been out there in Artarmon. We're up to level 3. Nothing to do with it. We're working with the gate investors on it, and there's some investors in that partnership, and we're giving them time that haven't invested in a development partnership before. So we're taking our time, doing it properly, making sure that the education is high. So there's been a number of visits out there. I think we're getting to the end of the diligence program. There was a document that went out a couple of days ago to the final piece of information everyone went, but nothing to do with the progress on the site is going very well. And just to be clear, we'll start marketing that, and we'll start to get serious about it beginning of next year. I don't know whether you've been following what's happening in North Sydney, but around Sydney generally, where it's getting harder to get planning and power and bigger gating. And I think Transgrid came out with a pretty good release today, which is going to make it more programmatic in regard to data center operators have got 5 or 6 gates to go through. We think that's all good. That's how we operate pretty well everywhere else in the world. You can imagine that a time where there's 90 megawatts of pretty prime data center space. So we're in a very, very good spot on that.

Got it. That's clear. Just on the logistics side of things. Look, you've been talking to activity there and project values increasing, I think, through the course of '26. It looks like logistics was fairly consistent with the third quarter and the half. So where could we kind of see that get to over the course of '27? Will it still be around that $4 billion mark or we think higher?

Yes, interesting. I was chatting to our Head of Industrial here in Australia just recently. And yes, look, Australia could be $3 billion by itself. So some of these -- so I wouldn't underestimate it. Some of these projects we're talking about with the automation are getting a real, real deliberate move by the customers. They're bigger buildings. They need 9, 10 megawatts of power, and they're fully roboticized. And we've got some of those going in Sydney. They're just bigger and they're more valuable. So I wouldn't underestimate the logistics phase over the next 4 or 5 years, we think is going to be pretty big.

That's great. Maybe just a last quick one. So San Jose, you had two sites there, I think you said were progressing well. Are they going to be an FY '27 story and still likely to be shells there?

Yes. It's going well, and it's -- what, are we in '26 at the moment? Yes, '27 will be good, yes. But there are a couple -- we'd have another -- we're working on 10 gigawatts of opportunity around the world, right? So you're going to find there's going to be other projects that come into it. We're going to move through some of these. The pipeline we've got is world-class under anyone's measure. And I'll just leave that with you. So we're not running done. We're serious about this. We think the hurdles around the world that are getting higher are good for us. We welcome it. And effectively, we're good for it. So planning, having the capital, you've got to own the land before you start having the conversation. Talk to your customers when you can demonstrate you're actually building something and you have what you say you have, then effectively being able to deliver on a coin, on a dime when they require it. That's the game we're in.

Our next question comes from Tom Bodor with Jarden.

I'd be interested in how much capital of your own capital and third-party capital sits in behind that $19.7 billion of WIP that's relating to data centers at the present moment? And also, where do you see that capital balance heading over the next, say, 3 years?

If I understand -- thanks, Tom, for the question. If I understand it correctly, the -- so 90% of those projects are already in third-party arrangements. And the construction of those is largely equity funded at the moment, and that's all understood and agreed equity finance. There's debt capacity already within those partnerships as well. Now depending on how much debt capacity we want to have at the end will be determined going forward. But for the moment, anyway, the work is largely equity funded and that's all pre-agreed. That's typically how we fund ourselves.

So how much capital is sitting behind the $19.7 billion of WIP, like actual dollar number today?

Well, if I give you that, I'm telling you the cost. So I'm not going to do that. But it's equity funded.

And I think, Nick, that's the point we went back, we made it early in the presentation. One of Goodman's big competitive advantages in the sector is actually, as it proved over time, this is not something new that we partner up, we spread the risk in the different return parameters. So it's development partnerships, which 90% of our development in those partnerships, Nick. So that means it spreads the risk across some of the biggest capital names in the world. And then effectively, the stuff we want to own and hold over time, we can alter how much we own. But fundamentally, those are also funding opportunities as well through holdcos, which we've been doing for a very long time. That is a tremendous competitive advantage in a world where capital is absolutely critical. And you guys follow the CapEx numbers, let's just say out of the top 4 or 5 hyperscalers in the U.S., I think the capital -- the CapEx number for this year is something like AUD 1 trillion, just to put in context, right? And when we also talk today about fully fitted, that's because what the market wants, right? So that will give you a sense of what they're trying to do by bringing credible third-party operators in rather than just all self-builds to actually be able to handle some of that load. So you need to be really good, really good at managing your capital and raising capital. Otherwise, you'll run out of runway in 5 seconds.

So I think to clarify the question, I meant how much capital today, not end cost.

Capital today. I'm not sure the difference between the way I answered the question. Maybe I just don't understand the question.

Okay. We'll chat about it later. The other one I was interested in is just your cadence of development starts since you raised in Feb last year, you've put 0.5 gigawatt into production. I was wondering how long it might take for the next 0.5 gigawatt to go to production.

20 billion, let's chew through that, right? I think everyone wants to see some leasing on the page. And we are adding anything to that number, you'll find there will be some customers in front of it. And then some of them will be bigger deployments, and we're working on some of those right at the moment. So look, let's get through what we've got on the page. This is a serious game we're in. Someone was chatting to me yesterday about the tortoise and the hare. Now I'm not saying we're the tortoise, but we don't want to be the hare either, right? We're going to do this properly. We're going to make sure we manage the capital properly, and we'll move through it when it's sensible and responsible.

I don't think anyone's ever characterized you as a tortoise, Greg.

No, it was a very good friend of mine who was making that comment.

And just a final one on the server side, the hyperscaler you've got into that 50 megawatts. Do you expect that same customer to deploy elsewhere globally in your portfolio? Or do they look at things on a more site-by-site localized basis?

Look, I won't talk about that customer specifically, but we're having conversations with customers across a number of countries with similar deployments. There's one actually a couple of nights ago. So yes, we are. We're very, very focused on good credit, good customers, because when Nick talks about holdco, right, and that's the capital that will own a lot of these prime data center assets over time. If you don't have that contract right and you don't have that set right, you can forget about holdco. That does not work. So we're super, super focused on quality. And if that means we take longer to lease upfront because we're more patient, so be it.

Our next question comes from Adam Calvetti with Bank of America.

Just a quick one. So it looks like there's about $8 billion in commencements this half if you kind of back solve DC proportion of WIP. That looks to be all data centers, and that is saying completed. That yield on cost of -- I think it's 9.1%. And considering 90% is fully fitted, seems lower than the double digit that you're quoting 6 to 12 months ago. Can you just break down what's going on?

The geography between Hong Kong, Japan, we're not going OTT on rents, but rents are moving quickly. So when we're looking at the programs at the moment on fully fitted in most places around the world where we're not talking about Japan and Hong Kong, like we separate that, we're certainly late 9s, in that 9 to 10, 11 range depending on the quality. But I come back to the quality of the comment I made earlier because the quality of what you put in the front end, it won't be any surprise to you, has got a direct correlation with the value at the back end, right? So there's no free lunch here. So what you're getting at the front is what you're going to produce at the back. So we've been relatively conservative on the -- what we put out and obviously, these documents. There's room in those. Our rents are actually moving at the moment, and we've got the costs pretty well locked down. So we'll see where it comes out, but we're in a very, very healthy state.

Look, I think we do talk ranges, right? And so a 20-year pre-lease to a global hyperscaler, you're not going to expect at the high end of the range, it's fair to say. Whereas if you're doing a colo facility with enterprise users, you would expect a significantly higher yield on cost. So we've talked about it on an average basis. And Greg is right. I mean what's typically in these numbers is a more conservative side of that estimate. But there is also -- I mean, we have started industrial projects as well, which are at lower yields. So that's why the average of the starts is where it is.

Okay. That's clear. And then just on the 9% EPS growth target, what leasing milestones, or data center leasing milestones, are embedded in that target? And then how do we think about the potential sell-down of Tokyo across the year? Is that embedded in the 9%?

Well, it's pretty important to note in Japan, we're not talking about Tokyo sell down. We're talking about building the building for customer, and Goodman Group will be owning that asset long term with partners just as we do at the moment, right? So we're not being specific about that in this. Effectively, though, I think running into 2027, I don't think we're -- we won't have a lot of assets completed. So I don't think there will be a lot of transfers around those. There might be some -- there could be some presales of some of these if we wanted to. But Nick and I will work that through with the teams around the world, what makes most sense as we go through the year.

Yes. I mean there's a lot of opportunities in front of us across sites that haven't even started yet right through to things that are in process.

We've got a few more partnerships we're going to be creating as well. So we'll just see how it balances out.

Okay. Just one quick one as well. Just how many sites outside of the 0.5 gigawatts have you started works on or site works on for DC use?

Just on the secured -- so if you go to the secured, right, because they are secured, we can -- we started early works packages on a lot of those things and moving earth around. We're moving earth around on a number of those at the moment. But if it's not secured, generally, we're not running around, moving earth around. Does that make sense?

Look, there's a couple of sites where there's some minor works going on, but nothing substantial.

Congrats on the result.

Our next question comes from Callum Bramah with Macquarie.

Just a couple. I think, Greg, you referred to a self-imposed limit on the amount of risk you take. Is there an actual quant number like a percentage of total assets that Goodman will have at risk in developments? And do you differentiate within that of speculative versus those that have a customer contract?

Good question. Nick, we were in a meeting about 3 days ago on that, weren't we?

Yes, yes. It was me, Cal. I'll line up on that one. I'll step up on that one. But the comment I made, you're right, and I'll talk to that in a second. The comment I made was more in relation to something we talked about a couple of years ago. We have a number of development opportunities. You remember, we parked some of them which were earmarked for industrial development. We parked them and said, hang on, there's power, there's opportunity here. And so we've diverted land resources as well as other and capital resources into data center development. And that's what I meant by self-imposed limitation, is we've diverted our resources. But as a general rule, there are risk limitations that we do work on, and they are sort of a combination of capital and earnings at risk. So there are limiters out there. We're working within those limiters at the moment, largely because we've entered into partnerships. And so that's given us capacity. So we do have capacity. I think it's more -- at the moment, what Greg talked about, it's more of a just commercial judgment and when is the right time and what's the right thing to do for each asset as we go.

So there's not a percentage of your total assets that you'll have exposed to development that you could share like 20%, 25% of total assets that is a cap?

No, we don't cap on the development asset portion. It's more dynamic and granular than that. We look at it on a kind of risk [ prioritization ] basis. So spec development, for example, is one of the areas that we look at. But you got to look at it in the context of where is our gearing, what's our earnings sensitivity to that? What's our liquidity look like? What are the actual risks? And so it's a bit more complex than that, too much detail to go through on this call, but the Board looks at it. We look at it every day. The Board looks at it every time we meet, the Risk and Compliance Committee looks at it. There's a range of different risk measurement tools that we use.