BlackRock IFM STACK Deal Nears $25 Billion as Power Becomes the Prize

BlackRock and IFM Investors have entered exclusive talks over STACK Infrastructure’s Asia Pacific data centers, with the portfolio reportedly valued at up to $25 billion.

The BlackRock IFM STACK deal remains a negotiation, not a completed acquisition. Bloomberg reported that the consortium was preparing to conduct due diligence, which allows a prospective buyer to test financial, technical, and legal assumptions.

That distinction matters because the headline value covers more than operating server space. The portfolio includes projects at different development stages across markets where land, electricity, permits, and construction capacity carry very different risks.

The contest is also larger than BlackRock versus another financial bidder. The real opponent is organic development, the slower alternative of securing land and power before building each campus from the ground up.

Buying STACK could give the consortium years of development work in one transaction. It could also transfer an expensive collection of unfinished obligations whose value depends on power arriving when customers need it.

The BlackRock IFM STACK Deal Has Reached Its Most Serious Stage

Exclusive negotiations give the consortium a privileged position, but they do not guarantee that STACK’s Asia Pacific portfolio will change hands.

According to Bloomberg’s account of the exclusive talks, the investor group includes BlackRock-backed Artificial Intelligence Infrastructure Partnership, or AIP, and IFM Investors. The group was preparing to examine the assets through due diligence.

Bloomberg placed the potential valuation between $20 billion and $25 billion, citing people familiar with private discussions. That range makes the proposed transaction important even within the increasingly large data center market.

The report did not identify a signed agreement, confirmed financing package, or completed regulatory process. Representatives for BlackRock, GIP, IFM, MGX, Blue Owl, and STACK had not publicly confirmed the transaction when the news emerged.

Due diligence will therefore carry more weight than the exclusivity headline. Buyers must confirm which facilities operate today, which are contracted, and which remain dependent on future construction.

They also need to establish the portfolio’s power position. A data center campus has limited value for AI customers if grid access, substations, cooling systems, or backup generation cannot support promised capacity.

Customer commitments will receive similar scrutiny. Long leases with established cloud companies can support financing, while speculative developments require assumptions about future demand and tenant quality.

The portfolio’s ownership history adds another layer. Blue Owl Capital acquired IPI Partners, the investment platform associated with STACK, after announcing the IPI acquisition in 2024.

Blue Owl said IPI’s broader platform covered 82 data centers and more than 2.2 gigawatts of leased capacity across several regions. Those figures described the wider investment platform, not the exact assets reportedly under discussion now.

That distinction prevents a common valuation mistake. STACK’s global statistics cannot automatically be assigned to the Asia Pacific sale perimeter.

STACK’s own materials describe operations across Singapore, Malaysia, Japan, South Korea, Australia, and other regional markets. However, the precise properties included in the proposed transaction remain unconfirmed.

The immediate change is therefore procedural but meaningful. A broad sale process has reportedly narrowed into exclusive talks with a named consortium.

That narrowing creates the article’s central tension. Buying an established development platform promises speed, but its reported valuation assumes the consortium can convert planned capacity into operating infrastructure.

Why BlackRock and IFM Want Capacity Instead of Another Construction Queue

The consortium is pursuing time, power access, and local execution capability, not simply buildings filled with servers.

Developing a hyperscale campus requires more than purchasing land. Operators must obtain permits, reserve electricity, connect transmission infrastructure, source equipment, arrange cooling, and satisfy local planning rules.

Each requirement can introduce years of delay. Constraints become harder when several operators seek the same industrial sites, contractors, substations, and grid connections.

STACK’s APAC portfolio illustrates why an existing platform appeals to institutional investors. The company says it has established teams and projects in several major regional markets.

Its Japan materials identify active projects in Inzai and Keihanna. They also describe more than 100 megawatts under development across five data centers in two Japanese markets.

Those company figures require cautious treatment. They describe STACK’s plans and operations, but they do not reveal contracted revenue, construction completion, or the transaction’s exact asset perimeter.

Still, they show what a buyer would be acquiring beyond concrete and electrical equipment. Local development teams, supplier relationships, permitting knowledge, and grid negotiations can compress expansion timelines.

This advantage explains why organic development is the BlackRock IFM STACK deal’s primary opponent. Starting independently might reduce the entry valuation, but it would expose the consortium to land and interconnection delays.

A transaction can consolidate those development risks inside one platform. It cannot make them disappear.

The attraction is especially strong for AIP. The partnership was established to mobilize long-term capital for AI infrastructure, including data centers and the energy systems supporting them.

AIP’s earlier Aligned Data Centers transaction demonstrates the strategy’s scale. In July 2026, AIP, MGX, and BlackRock’s Global Infrastructure Partners completed the Aligned acquisition.

Aligned said that portfolio covered 51 campuses and more than 6.4 gigawatts of operating and planned capacity. The consortium also committed additional growth capital for expansion.

That transaction concentrated on North and Latin American markets. STACK would give an AIP-backed group a substantial path into Asia Pacific, assuming the reported talks become an agreement.

IFM adds a complementary infrastructure perspective. Its investment model emphasizes long-duration assets, while its portfolio experience spans transportation, utilities, and digital infrastructure.

Data centers sit at the intersection of those categories. Their commercial performance depends on property, electricity, cooling, telecommunications, financing, and long customer contracts.

Buying the platform could let the consortium coordinate those inputs at regional scale. Building a comparable footprint project by project would require many separate negotiations.

Speed has become particularly valuable as customers prepare AI systems that use higher-density computing. Those deployments require carefully designed power and cooling systems, which cannot be improvised after a building opens.

The consortium is effectively deciding whether STACK’s development head start justifies the reported price. Due diligence must determine how much of that head start is real.

Power Access Is Driving Asia’s Data Center Valuations

Electricity availability now separates strategically valuable campuses from impressive development maps that cannot be delivered on schedule.

CBRE reported that Asia Pacific data center investment reached a record $11.6 billion in 2025. Entity-level transactions accounted for $8.3 billion during the same year.

The firm’s regional investment analysis said AI demand was redirecting growth toward power-advantaged markets. Malaysia, Australia, and India were among the locations benefiting from that shift.

Johor’s live capacity increased 53 percent during 2025, according to CBRE. Melbourne recorded 37 percent growth, while Singapore and Hong Kong grew between 6 percent and 8 percent.

These figures reveal a market split. Established hubs retain customers and network connectivity, but emerging locations can offer more room for large, high-density campuses.

Singapore shows the tension clearly. The market offers strong connectivity and customer demand, yet new projects face tightly managed electricity and land constraints.

South Korea presents another constraint. CBRE said authorities had restricted new projects within Greater Seoul to 10 megawatts, encouraging developers to consider other locations.

Japan combines deep enterprise demand with difficult construction conditions and lengthy power planning. Australia offers land and renewable energy potential, but its largest campuses still depend on transmission and generation schedules.

Malaysia, especially Johor, has gained attention because it combines available land with proximity to Singapore. However, water, grid readiness, and policy changes remain material planning variables.

A regional portfolio can balance these differences. Capacity in one market may serve customers whose preferred location cannot accept another large project.

That flexibility matters for hyperscalers, which are the largest cloud platforms. They often reserve capacity years before a facility begins operating.

It also matters for neoclouds, specialized providers that rent AI computing infrastructure. Their demand can be substantial, although landlords may view their credit quality more cautiously.

A buyer must therefore examine both technical capacity and revenue quality. A megawatt reserved by an established customer does not carry the same risk as speculative capacity.

The valuation also depends on power milestones. A site with a signed interconnection agreement differs from land where electricity remains an ambition.

Infrastructure investors usually separate operating assets from projects under construction and longer-term development pipelines. Each category receives different assumptions about timing, cost, and risk.

The reported BlackRock IFM STACK deal compresses those categories into one headline number. Due diligence must unpack them again.

This is why the deal is not merely a bet that AI demand will grow. It is a bet that specific campuses can secure energy, customers, permits, and equipment before competing projects.

That mechanism places utilities and regulators inside the competitive landscape. They can influence which operators deliver capacity, even when they never bid for an asset.

Buying STACK Would Extend a Global Infrastructure Land Grab

Large asset managers increasingly treat data center platforms as strategic operating systems for capital, rather than collections of isolated properties.

A data center platform can repeatedly deploy capital across multiple campuses. It maintains design standards, procurement relationships, customer contacts, and development teams.

Those capabilities make platform acquisitions attractive to investors seeking exposure beyond a single facility. They also create potential economies across financing and equipment purchasing.

The Aligned transaction established a recent precedent for BlackRock’s infrastructure strategy. It paired large institutional capital with an operator that could continue developing campuses under existing management.

If completed, the STACK transaction would apply a similar model in another region. The buyer would obtain a development organization alongside physical and planned assets.

Blackstone and CPP Investments used comparable logic when they agreed to acquire AirTrunk in 2024. The AirTrunk transaction covered a platform operating across Australia, Japan, Malaysia, Hong Kong, and Singapore.

AirTrunk had more than 800 megawatts committed to customers, according to Blackstone. It also controlled land that could support more than one gigawatt of future development.

That deal provided an important regional benchmark, but it should not be treated as a direct valuation template. Portfolio maturity, customer commitments, financing, and project definitions can differ widely.

STACK competes within the same broad race for hyperscale demand. AirTrunk, Equinix, Digital Realty, Keppel, NEXTDC, and other operators pursue overlapping customers across parts of Asia Pacific.

Their strategies vary. Some emphasize retail colocation, where many customers rent smaller deployments. Others focus on wholesale campuses designed for a few large cloud or AI tenants.

STACK presents itself as a provider of large campus, build-to-suit, colocation, and powered-shell solutions. A powered shell supplies the building and electrical infrastructure while leaving some internal equipment to the tenant.

This range gives customers several ways to secure capacity. It also exposes an operator to different construction schedules, contract structures, and capital requirements.

The BlackRock-backed group would gain a platform positioned across those models. Yet rivals would retain meaningful advantages in specific markets.

Established operators may already hold customer relationships, operational facilities, and interconnection rights. Local developers may move faster through municipal approval processes.

The competitive response is unlikely to arrive as one rival acquisition. It will appear through leases, land purchases, partnerships, and accelerated construction announcements.

Large technology customers also possess negotiating power. They can divide workloads among operators or delay commitments if proposed lease terms become unattractive.

That limits the pricing freedom of even a scarce portfolio. Power access creates value, but customers still assess reliability, location, latency, sustainability, and total cost.

The acquisition would therefore expand BlackRock and IFM’s regional position without settling the market. Execution after closing would remain more important than ownership on closing day.

The Reported Valuation Still Faces a Difficult Reality Check

The greatest risk is that buyers pay today for capacity whose electricity, customers, or completion dates remain uncertain.

The proposed transaction is based on anonymous-source reporting. No public definitive agreement currently establishes the final valuation, financing, included assets, or closing schedule.

Exclusive talks can end without a sale. Buyers can seek revised terms after discovering higher costs, delayed projects, weaker contracts, or regulatory complications.

The difference between the earlier reported sale expectations and the current range deserves attention. It might reflect negotiations, a changed asset perimeter, financing conditions, or updated project assumptions.

Public reporting does not yet provide enough evidence to choose among those explanations. Treating the difference as a confirmed price reduction would overstate what is known.

Asset composition presents another uncertainty. STACK publishes global and country-level capacity figures, but those numbers combine operating, developing, planned, and potential projects.

Potential development is not equivalent to operating capacity. It can depend on land options, customer demand, government approvals, and future grid investment.

Construction costs can also shift during a long development cycle. Electrical equipment, transformers, generators, cooling systems, and skilled labor all affect the final capital requirement.

Financing adds another variable. Data center acquisitions often combine equity with significant debt, and borrowing terms influence returns even when customer demand remains strong.

Currency exposure matters across a regional portfolio. Construction spending and revenue may occur in Australian dollars, yen, won, ringgit, or Singapore dollars.

Regulatory reviews could examine foreign ownership, national security, data sovereignty, or critical infrastructure. Requirements differ by jurisdiction and can change deal timing.

Customers may also reconsider deployments. AI demand is expanding, but model efficiency, chip supply, and corporate spending decisions influence how quickly reserved computing capacity becomes necessary.

The strongest demand forecasts cannot substitute for signed leases. Buyers need to distinguish genuine customer commitments from expressions of interest.

Environmental constraints add pressure. Large facilities require electricity and cooling, while communities increasingly scrutinize water use, grid impacts, emissions, and local economic benefits.

Sustainability claims need equal scrutiny. Renewable procurement does not always mean a campus receives carbon-free electricity during every hour of operation.

The BlackRock IFM STACK deal also creates concentration risk. One transaction would place a large regional development pipeline under a consortium already making major digital infrastructure commitments.

Scale can improve procurement and financing. It can also magnify delays if projects share the same equipment suppliers, customers, or power bottlenecks.

The cautious conclusion is not that the valuation is unjustified. It is that the public evidence cannot yet demonstrate whether the reported range matches the portfolio’s deliverable capacity.

Readers should resist dividing a headline valuation by a global capacity figure. That calculation would mix assets, regions, development stages, and uncertain future projects.

A credible valuation assessment requires the final asset list, operating earnings, contracted capacity, remaining construction spending, and power-delivery milestones. None has been publicly disclosed for this transaction.

What to Watch Before the Deal Becomes an AI Infrastructure Test

Three signals will determine whether exclusive talks become a defensible acquisition or another ambitious data center proposal.

The first signal is a definitive agreement with a precise transaction perimeter. That announcement should identify the buyer group, seller, included markets, and expected closing conditions.

A signed agreement would strengthen the view that the consortium has accepted STACK’s development risks after reviewing confidential information. Continued silence would weaken that interpretation.

The announcement may also clarify whether MGX, GIP, or other AIP participants will provide capital. Their roles cannot be assumed from earlier transactions.

The second signal is verified operating and contracted capacity. Investors need separate figures for live facilities, construction projects, committed customers, and longer-term development land.

Those disclosures would show whether the valuation rests primarily on current cash flow or future delivery. A portfolio weighted toward contracted capacity would support the speed argument.

A portfolio dominated by prospective projects would shift attention toward construction spending and execution. It would make the purchase resemble a development bet rather than an operating-asset acquisition.

The third signal is regulatory and power progress across the largest campuses. Watch for foreign-investment approvals, environmental permits, grid agreements, and construction milestones.

Timely approvals and power delivery would reinforce the consortium’s decision to buy an existing platform. Material delays would strengthen the case for a lower valuation or staged investment.

These signals matter beyond infrastructure finance. Cloud customers, AI developers, and enterprise technology buyers all depend on the capacity that operators can actually deliver.

More regional supply can create additional deployment options and reduce dependence on a few constrained hubs. Delayed power connections can produce the opposite result.

Enterprise buyers should also watch how ownership affects contracts and expansion plans. A capital-rich consortium may accelerate projects, but commercial priorities can change after an acquisition.

Knowledge workers will not choose data center sites directly. Still, the availability and cost of computing infrastructure influence the AI services they use every day.

Teams following these announcements can organize filings, project updates, and vendor claims inside a searchable AI knowledge base. The goal is to separate confirmed milestones from repeated projections.

The BlackRock IFM STACK deal is therefore best understood as a contest between purchased speed and development reality. Exclusive talks establish momentum, not completion.

The next announcement should answer a simple question: did the consortium buy operating capacity, a credible power pipeline, or an expensive promise to build both?