Goodman’s A$19.7bn development book is now 78% data centres as AI demand reshapes the group

Goodman Group (ASX: GMG) has entered FY27 with data centres accounting for 78% of its A$19.7 billion development work in progress, turning what was historically a global logistics property group into an increasingly important provider of digital infrastructure. FY26 operating profit increased 15.7% to A$2.675 billion and operating earnings per security rose 10.1% to 129.9 cents, while statutory profit reached A$2.779 billion. Development earnings grew even faster at 34% to A$1.792 billion as Goodman accelerated construction across power-constrained metropolitan markets, with more than A$15 billion of current development work now tied to data centres. Management is targeting another 9% increase in operating EPS in FY27, but the scale of the construction program means future earnings increasingly depend on converting secured power and partially committed developments into leased, completed and ultimately income-producing infrastructure.

That transition is occurring much faster than the headline growth of Goodman’s A$89 billion property portfolio might suggest. Development work in progress expanded by more than 50% during FY26, the data-centre power bank increased from 5.0GW to 6.4GW, and approximately 0.5GW is already under construction through ten projects in eight major cities. Yet only around half of projects currently in work in progress are either leased or in advanced negotiations, which means Goodman is deliberately progressing construction alongside customer commitments rather than waiting for every development to be fully pre-leased. The strategy can improve speed to market in cities where power and suitable land are scarce, but it also makes leasing conversion, construction execution and partnership funding more important determinants of FY27 and FY28 returns.

Goodman securities closed August 20 at A$28.78, down 1.51% after trading between A$28.07 and A$29.75. The security is approximately 4.2% below its August 13 close of A$30.05 and about 1.3% below its July 20 close of A$29.16, while remaining roughly 23% below the A$37.31 52-week high and 17% above the A$24.56 low. The subdued result-day reaction suggests investors are balancing another strong operating result against the execution requirements embedded in a development workbook whose scale has moved well beyond Goodman’s traditional logistics development cycle.

How quickly has Goodman Group’s data-centre exposure changed the development business?

Goodman’s A$19.7 billion development work in progress spans 50 projects across 12 countries and carries an expected yield on cost of 8.2%. Data centres now represent 78% of that total, implying more than A$15 billion of active development exposure to the sector, compared with 73% of A$14.4 billion of work in progress at the December 2025 half-year. The absolute data-centre development book has therefore expanded substantially in only six months, while total WIP has moved well beyond management’s earlier expectation that it would approach A$18 billion by June.

The earnings contribution is already visible. Development earnings increased 34% to A$1.792 billion and were the largest contributor to Goodman’s FY26 result, while the group commenced A$8.1 billion of new developments and completed A$3.6 billion during the year. The annualised production rate has risen above A$7.5 billion, indicating that the current A$19.7 billion workbook is not simply a long-dated land pipeline but is increasingly being converted into active construction.

The composition distinguishes this cycle from Goodman’s earlier logistics growth. Modern warehouses can require substantial land and infrastructure, but fully fitted hyperscale data centres add far greater electrical, mechanical and cooling requirements and depend critically on securing large amounts of grid capacity. Goodman is therefore committing more capital and managing more technically complex developments, although the expected returns are also potentially higher where power scarcity limits competing supply.

The strategic payoff becomes much larger if Goodman retains exposure beyond development completion. Management says the majority of data-centre projects presently under construction are fully fitted developments and expects to operate some facilities for customers, potentially allowing the group to capture investment and management income after development profits have been recognised. Delivery of the current projects is staggered between early 2027 and 2030, so the earnings effect should extend well beyond a single reporting period.

How much of Goodman’s A$19.7 billion development pipeline is actually committed by customers?

Approximately 50% of projects in work in progress are either leased or in advanced negotiations, according to Goodman. The number is encouraging given how rapidly the development program has expanded, but it also means a meaningful portion of current WIP is progressing before binding leasing outcomes have been disclosed. Management’s strategy is to construct into anticipated demand in supply-constrained markets where customers may otherwise be unable to obtain facilities when required.

There is already evidence of large customers committing to the model. Goodman recently signed a 20-year lease covering 50MW with a hyperscale customer in Tokyo and says whole-building discussions are at advanced stages across several other locations. The company argues that cloud expansion and the shift from centralised AI training toward inference closer to end users are increasing demand for metropolitan data-centre capacity, particularly where new electricity connections and development land are difficult to obtain.

The timing of customer agreements matters because data-centre projects can consume substantial capital before rent begins. A project that secures power but waits years for a customer commitment has fundamentally different economics from one where construction progresses alongside a long-duration hyperscale lease. Goodman’s FY27 execution therefore needs to be judged not only by how quickly WIP expands but by how much of that workbook moves from prospective demand into contracted utilisation.

The company’s A$3.6 billion of FY26 development completions were 89% leased, which provides some evidence that speculative exposure has historically converted into customer demand by completion. Maintaining a comparable leasing outcome as data centres become a much larger proportion of development activity would strengthen the case that Goodman is managing the higher construction exposure without sacrificing occupancy discipline.

Why is Goodman’s 6.4GW power bank potentially more important than its current property portfolio?

Goodman’s global power bank increased to 6.4GW across 16 major cities from 5.0GW across 13 cities a year earlier. Of that total, 3.6GW is secured and another 2.8GW is in advanced stages of procurement, while only approximately 0.5GW has so far entered active development. The gap between 6.4GW of potential power capacity and 0.5GW under construction gives Goodman a much larger future development option than the current WIP number alone captures.

Power has become one of the principal constraints on global data-centre development because AI and cloud infrastructure can require hundreds of megawatts at individual campuses. Acquiring suitable land without an achievable grid connection is therefore insufficient, while already having secured power in major urban markets can materially shorten development timelines and improve a site’s strategic value. Goodman has spent several years assembling both land and electrical capacity, which is why management increasingly describes the business as a provider of digital infrastructure rather than solely an industrial landlord.

The conversion rate will be more important than the headline power-bank figure. If a meaningful proportion of the remaining secured capacity moves into WIP over the next several years, Goodman’s current A$19.7 billion workbook could remain elevated even as existing projects reach completion. Conversely, secured power that does not attract suitable customers or produces returns below the group’s development hurdle would be less valuable than the headline gigawatt figure suggests.

The current construction program represents fewer than one-tenth of the 6.4GW power bank, illustrating the duration of the opportunity but also the amount of execution still required. This makes power procurement a leading indicator of future potential rather than the equivalent of contracted development revenue.

Can Goodman finance a data-centre construction cycle this large without materially increasing leverage?

Goodman’s group gearing increased from 4.3% to 6.5% during FY26 as development activity accelerated, while look-through gearing reached 19.5%. Both measures remain conservative relative to the scale of the development program, and the group finished June with A$6.4 billion of cash and undrawn lines. Interest cover stood at 25.4 times at the group level and 9.5 times on a look-through basis.

The partnership model is critical to keeping leverage comparatively low. Seventy-one percent of current WIP is being undertaken for Partnerships or third parties, while approximately 90% of data-centre developments under construction, including associated expansion land, sits within Partnerships. This structure allows Goodman to retain economic exposure through co-investment, development earnings and management fees without funding the entire capital requirement from its listed balance sheet.

The funding pool has also expanded alongside the workbook. Goodman raised approximately A$3.2 billion of third-party equity during FY26 and created four new capital Partnerships, taking the total to 26. External assets under management increased 5% to A$75.4 billion, while the Partnership platform had A$6.7 billion of cash and undrawn debt plus A$5.7 billion of conditional equity commitments at year-end.

An Australian development Partnership is expected to be established during the first half of FY27, adding another potential funding vehicle as the data-centre program scales. This financing architecture is a major reason Goodman can contemplate a A$19.7 billion workbook while retaining single-digit group gearing, although the model remains dependent on institutional investors continuing to allocate substantial capital to data-centre and logistics strategies.

Why does Goodman retain so much profit instead of paying a larger distribution?

Goodman generated operating earnings of 129.9 cents per security in FY26 but distributed only 30 cents, leaving the majority of operating earnings inside the business. The distribution therefore represents only about 23% of operating EPS, which is unusually low compared with many traditional listed property vehicles but is consistent with Goodman’s development-heavy strategy.

The retained capital is particularly valuable during the current data-centre cycle because Goodman has more development opportunities than could realistically be funded while paying out most annual earnings. Management has repeatedly prioritised reinvesting cash into developments and investment partnerships where it believes returns exceed the value of increasing near-term distributions. The 8.2% forecast yield on cost across current WIP provides one measure of the economics management is targeting, although actual development outcomes can differ from projected yields.

This policy changes how Goodman should be compared with higher-yielding A-REITs. Securityholders are accepting a relatively modest cash distribution in exchange for management retaining capital to compound the development and investment-management businesses. That trade-off works when development returns remain attractive and operating EPS grows, but it places greater emphasis on capital allocation because retained earnings represent cash that could otherwise have been distributed.

FY26 operating EPS growth of 10.1% exceeded the company’s original 9% target, providing support for the reinvestment strategy. Goodman is again targeting 9% OEPS growth for FY27, so continued double-digit or high-single-digit earnings progression will be important in demonstrating that the capital retained from shareholders is still being deployed productively.

Is Goodman still a logistics property company when development earnings dominate the result?

Goodman’s existing property portfolio remains a large and resilient earnings foundation. The total portfolio increased 4% to A$89 billion, occupancy remained high at 95.6%, like-for-like net property income grew 4%, and management estimates current rents sit approximately 10% below market levels on average. Property investment income increased 7% to A$722.1 million, supported by rental growth and higher average capital invested.

Management earnings provide another recurring component. External AUM increased to A$75.4 billion and management earnings reached A$690.1 million, supported by higher base fees and property-services income, although transactional and performance fees were lower. These established businesses give Goodman recurring income streams that reduce its dependence on any single year’s development completions.

Development earnings of A$1.792 billion are nevertheless now much larger than either property investment income or management earnings individually. The 34% increase in development earnings was also substantially faster than the 7% growth in property investment income, demonstrating where incremental group earnings are currently being created.

Goodman has therefore not stopped being a logistics landlord, but its growth profile is increasingly determined by development and particularly data centres. The existing logistics portfolio provides locations, customer relationships, recurring rental income and access to capital partners, while digital infrastructure is becoming the principal accelerator of earnings and capital deployment.

How much risk is Goodman taking by building data centres before all projects are leased?

The strategy carries more development exposure than waiting for full pre-commitments, but Goodman is attempting to mitigate that risk through location scarcity, secured power, partnership capital and staged construction. Approximately half of WIP is already leased or in advanced negotiations, while the majority of active data-centre construction is held through Partnerships rather than solely on Goodman’s balance sheet.

The company’s rationale is that hyperscale customers increasingly require speed and certainty. If Goodman waited for a final lease before beginning every project, customers could face multi-year construction and grid-connection delays, reducing the commercial value of the sites. Starting earlier allows Goodman to offer delivery windows when competing sites may still be waiting for power or approvals.

The risk becomes greater if AI infrastructure spending slows materially or customers alter their computing architectures before projects complete. Data centres are highly specialised assets, and a fully fitted facility carries more customer and technology exposure than a conventional warehouse shell. Delivery dates extending from early 2027 through 2030 mean the current workbook spans several years in which customer demand, hardware density and cooling requirements could continue evolving.

Goodman’s protection lies partly in metropolitan land scarcity and the adaptability of sites that can often support multiple phases of development. The company also retains a large logistics business, meaning suitable industrial land can in some cases have alternative uses, although that flexibility will vary materially by site and by how much data-centre-specific infrastructure has already been installed.

What does a 9% FY27 operating EPS target imply after FY26 beat guidance?

Goodman finished FY26 with operating EPS of 129.9 cents, up 10.1%, compared with the 9% growth target management had maintained throughout the year. Applying the new 9% FY27 target mechanically to FY26 OEPS would imply approximately 141.6 cents per security, although that figure is a BNT calculation rather than explicit company guidance for cents per security.

The target appears relatively measured when compared with the 34% growth in FY26 development earnings and the greater than 50% expansion in WIP. That difference likely reflects the timing between construction activity and profit recognition, as well as the contribution of property investment, management income, financing costs and corporate expenses across the broader group.

It also gives Goodman some room to absorb execution variability as the data-centre program scales. Developments can move between reporting periods because of leasing, completion, capital-partner transactions and valuation timing, meaning a very large workbook does not translate evenly into operating EPS every year.

The more important measure may be whether Goodman can sustain high-single-digit or better operating EPS growth for several years while converting the data-centre pipeline. Deliveries from the current workbook stretch through 2030, while the 6.4GW power bank provides potential development capacity beyond the projects already under construction, creating a longer runway than a single FY27 target captures.

Key takeaways from Goodman Group’s FY26 result and data-centre expansion

• Goodman delivered FY26 operating profit of A$2.675 billion, up 15.7%, while operating EPS increased 10.1% to 129.9 cents and statutory profit reached A$2.779 billion.
• Development work in progress reached A$19.7 billion across 50 projects in 12 countries, representing growth of more than 50% during FY26.
• Data centres account for 78% of WIP, putting more than A$15 billion of the current development workbook into digital infrastructure.
• Goodman has approximately 0.5GW of data-centre developments underway through ten projects in eight global cities, with current deliveries staged between early 2027 and 2030.
• The global power bank has increased to 6.4GW, comprising 3.6GW of secured power and another 2.8GW in advanced procurement.
• Approximately 50% of current WIP is leased or in advanced negotiations, while a recently signed Tokyo agreement covers 50MW for 20 years with a hyperscale customer.
• Development earnings increased 34% to A$1.792 billion, compared with property investment income of A$722.1 million and management earnings of A$690.1 million.
• Group gearing increased from 4.3% to 6.5% but remains low, while Goodman has A$6.4 billion of group cash and undrawn facilities.
• External AUM increased 5% to A$75.4 billion, with approximately A$3.2 billion of third-party equity initiatives completed during FY26.
• Goodman is targeting another 9% increase in operating EPS in FY27 after exceeding its FY26 growth target.

What must Goodman prove as data centres become the dominant source of development growth?

Goodman’s FY26 result provides strong evidence that the data-centre strategy has moved beyond land banking and power procurement into material earnings generation. Development earnings increased 34%, the workbook expanded above A$19 billion and 0.5GW of capacity is now under construction, while the broader logistics portfolio continues producing high occupancy and positive rental growth. The combination gives Goodman a recurring property and management earnings base beneath a much faster-growing development engine, which helps explain why the company can pursue digital infrastructure without abandoning the operating resilience of its established industrial portfolio.

The next stage carries a different execution challenge because the development workbook has become sufficiently large that further WIP growth alone will provide diminishing evidence of success. Goodman needs customer commitments to progress alongside construction, projects to reach completion at attractive returns and the partnership model to continue supplying external capital without forcing a substantial increase in group leverage. With only around half of current WIP leased or in advanced negotiations and deliveries extending through 2030, leasing conversion will increasingly determine whether the current A$19.7 billion pipeline becomes durable investment and management income after development profits are recognised.

The funding position provides meaningful protection. Group gearing remains only 6.5%, partnerships carry substantial cash, undrawn debt and conditional equity commitments, and approximately 90% of data-centre construction plus associated expansion land is already held through partnership structures. That gives Goodman flexibility to keep developing into constrained markets while sharing much of the capital requirement with institutional investors, although it also makes continued demand from those investors an important component of the growth model.

FY27 will therefore be less about proving that demand for AI and cloud infrastructure exists and more about demonstrating that Goodman can monetise its scarce land and power position without allowing construction risk to outrun customer commitments. Another 9% increase in operating EPS would extend an already strong earnings record, but the more consequential proof will come from the proportion of the A$15 billion-plus data-centre workbook that becomes leased, completed and ultimately retained within income-generating partnerships. If that conversion progresses while gearing remains low, Goodman will have stronger evidence that its shift toward digital infrastructure is not simply producing a larger development pipeline but a structurally larger earnings platform.