{"id":14406,"date":"2026-03-16T06:02:32","date_gmt":"2026-03-16T11:02:32","guid":{"rendered":"https:\/\/arc-group.com\/?p=14406"},"modified":"2026-03-17T02:44:15","modified_gmt":"2026-03-17T07:44:15","slug":"southeast-asia-data-centre-ma-2026","status":"publish","type":"post","link":"https:\/\/arc-group.com\/southeast-asia-data-centre-ma-2026\/","title":{"rendered":"Southeast Asia Data Centre M&A in 2026: Why AI Is Driving the Next Wave of Mega-Deals"},"content":{"rendered":"

Executive Summary<\/span><\/h2>\n

Southeast Asia’s data centre market is at an inflection point. AI workloads, sovereign cloud mandates, and hyperscaler expansion are driving demand at an unprecedented pace \u2014 while supply remains constrained by power limitations, permitting delays, and tightening regulatory requirements. This dislocation is catalysing a new wave of M&A, with 2026 marking a significant acceleration in that trend.<\/p>\n

In this report, we identify three dominant M&A patterns<\/strong> shaping 2026 dealmaking: platform acquisitions outpacing single-asset trades, minority-to-control structures emerging as the default entry mechanism, and telco carve-outs unlocking embedded value. Our central thesis is that platform-level assets, those combining secured power, multi-market execution capability, and a proven delivery track record \u2014 will command valuation premiums of 25\u201335x EBITDA<\/strong>, a material step-up from single-asset trades, as supply constraints make replicable development pipelines the scarcest commodity in the market.<\/p>\n

With regional data centre capacity projected to triple by 2030 and the generative AI market growing at a 50% CAGR to reach US$13B, the demand fundamentals are undeniable. Yet M&A has emerged as the fastest and most effective route to scale precisely because securing land, power, and permitting from scratch is no longer a viable path for most buyers. For a detailed overview of where growth is concentrating across the region, see Harnessing ASEAN’s Data Center Boom.<\/a> For a deeper look at how investors and operators can position themselves across these dynamics, see Strategic Positioning for Data Center Investment in Southeast Asia<\/a>.<\/p>\n

I. Key Trends: Why 2026 Marks an Inflection Point<\/span><\/h2>\n

AI and sovereignty are the new demand drivers<\/h3>\n

AI workloads require far greater power density than traditional cloud computing. This is not a linear growth story but an exponential one. Southeast Asia’s data centre capacity is projected to triple by 2030 compared to 2025 levels and reach a range of 5.2 GW to 6.5 GW driven by a tenfold increase in AI computing as well as demand for traditional computing. The generative AI market in Southeast Asia is forecast to grow at a 50% CAGR, from US$0.8B in 2023 to US$13B by 2030.<\/p>\n

This surge is driving a fundamental shift in data centre requirements. Generative AI models, particularly large language models (LLMs), require immense computational power for both training and inferencing<\/strong>. Training demands massive, concentrated computing power, pushing data centre power demand in the Asia-Pacific region up by an estimated 165% by 2030<\/strong> compared to 2023, necessitating high-density server clusters and advanced cooling. Inferencing\u2014deploying trained models for real-time applications\u2014demands low-latency facilities near urban hubs, reshaping location strategies and fuelling growth in edge data centres.<\/p>\n

Simultaneously, sovereign cloud initiatives<\/strong> are gaining traction. Malaysia’s 2026 budget allocated MYR (~US$490M) for a sovereign AI cloud, aiming to keep data and model training within national borders. This allocation is part of a wider commitment totalling MYR5.9B (~US$1.44B) for research, development, commercialisation, and innovation initiatives relating to AI. Combined, the roadmap and commitment bolster the country’s vision of AI sovereignty: achieving domestic control over AI development to accelerate economic growth.<\/p>\n

Other countries in the region are also advancing sovereign digital agendas. In Indonesia, the cloud computing market has experienced remarkable growth, with a CAGR of 48% over the last 5 years, surpassing the global average. Currently, 90% of companies in Indonesia are moving towards cloud computing solutions, driven in part by government and enterprise adoption. Meanwhile, Singapore migrated most of its less sensitive digital workloads to the commercial cloud, demonstrating the maturity of its cloud-first policies. The scale of sovereign digital ambition across the region is further underscored by the capital commitments of hyperscalers, who have backed these markets with some of the largest infrastructure investment pledges in the industry’s history. AWS committed SG$12B (~US$9B) to Singapore between 2024 and 2028 and launched its first Thai cloud region in January 2025. Google has invested a cumulative US$5B in Singapore and pledged a further US$2B to develop its first Malaysian data centre and cloud region. Microsoft followed with US$2.2B earmarked for Malaysia and US$1.7B for Indonesia, both over the 2024\u20132028 period. Taken together, these commitments exceed US$20B in announced capex across the region which translates directly into land acquisition, power procurement, and construction pipelines that will define M&A supply for years to come.<\/p>\n

The ASEAN Framework on Cross-border Cloud Computing<\/strong>, endorsed in January 2026, operationalizes the concept of “Trusted Data Corridors” (TDCs). This mechanism allows for the unimpeded flow of data between accredited data centres in participating member states, provided their domestic data protection regimes are deemed legally comparable. The initial corridors are expected to form between “first mover” nations with mature digital economies, such as Singapore, Malaysia, and Indonesia. For these countries, the TDC framework serves a dual purpose: it lowers compliance costs and technical barriers for regional platform consolidation, while simultaneously creating a strong economic incentive for tech giants and local enterprises to host data within ASEAN by designating in-region data centres as “trusted”. This effectively fosters an environment where data is kept within ASEAN’s geopolitical boundaries but can still flow freely within the bloc’s secure digital zone.\"Map<\/a><\/p>\n

As illustrated above, a TDC connects designated data centres across two ASEAN Member States (AMS) where one state (AMS A) joins via its entire territory and another (AMS B) via a defined Special Economic Zone, linked through a government-to-government agreement. Critically, the two states need not sit within the same cloud Availability Zone (AZ) \u2014 the logical groupings of physical data centres that cloud providers use to deliver services across a region. The framework is designed to bridge across AZs, as shown by AMS A in AZ 1 and AMS B’s SEZ straddling AZ 2. Within this corridor, data moves freely between accredited facilities without triggering the cross-border transfer restrictions that ordinarily apply under each state’s domestic data protection laws \u2014 effectively creating a single regulatory zone for data flows, while preserving each state’s sovereignty outside the corridor’s boundaries.<\/p>\n

Supply-side paralysis<\/h3>\n

While demand increases, new data centre supply faces three structural barriers across Southeast Asia: government-imposed moratoria on non-AI facilities, tightly controlled capacity allocations with stringent green energy requirements, and permitting and grid access delays.<\/p>\n

Malaysia<\/strong> has restricted new data centre developments not related to artificial intelligence for the past 18\u201324 months to manage power and water consumption. Prime Minister Anwar Ibrahim confirmed in February 2026 that “all new applications that were not related to AI have already been stopped,” with approvals only granted to projects offering high-technology benefits related to AI. In Johor, state authorities have imposed tighter requirements on water and power use for new data centres to prevent strain on local resources.<\/p>\n

Singapore<\/strong>, despite being the region’s Tier-1 hub with 1GW operational capacity and a 1.4% vacancy rate, maintains strict control over new supply. The government’s second Data Centre Call for Application (DC-CFA2), closing March 31, 2026, offers at least 200 MW of new capacity\u2014but only to operators meeting specific requirements: Power Usage Effectiveness (PUE) below 1.25, Green Mark Platinum certification, and at least 50% green energy sourcing from approved renewable sources. Land scarcity, with Singapore’s total area of 734 square kilometres, imposes fundamental limits on expansion.<\/p>\n

Other regional markets<\/strong> face their own structural barriers. Indonesia’s project timelines are slowed by unreliable coal-heavy power, slow renewable energy approvals, grid access uncertainties, and drawn-out permit procedures. Thailand, despite strong demand, faces challenges in delivering power to locations where operators want to build. Vietnam has high power availability and demand, but its permitting system remains a challenge, with investors waiting for successful builds to demonstrate the system works.\"Map<\/a><\/p>\n

The PUE Factor: Efficiency as a Barrier to Entry<\/h3>\n

As power becomes the primary constraint, how efficiently a data centre uses that power has become a critical differentiator. Power Usage Effectiveness (PUE)\u2014the ratio of total facility energy to IT equipment energy\u2014is no longer just an environmental metric; it is a competitive and regulatory gatekeeper.<\/p>\n

Singapore’s Green Data Centre Roadmap requires new builds to achieve a PUE below 1.3, the most stringent standard in Asia-Pacific. This means operators must invest in advanced cooling technologies, liquid cooling systems, and smarter facility design from day one. For existing assets, a sub-1.3 PUE commands a valuation premium because it represents lower operating costs, regulatory compliance, and future-proofing against carbon taxes.<\/p>\n

For acquirers, underwriting PUE is now as important as underwriting power availability. A platform with consistently low PUE across its portfolio signals technical expertise and operational discipline\u2014capabilities that are hard to replicate and increasingly scarce. In a market where power is rationed, the operators who do more with each megawatt will be the ones who win.<\/p>\n

II. The Industry Value Chain: A Systematic Overview<\/span><\/h2>\n

To understand where M&A activity will concentrate, we must first map the industry structure and the key forces shaping it. The Southeast Asia data centre ecosystem can be visualized as a layered value :\"Chart<\/a>The value chain spans five layers: IT infrastructure and OEMs (servers, semiconductors, and hardware), non-IT infrastructure and contractors (cooling, power distribution, and construction), data centre operators (colocation and wholesale facilities), hyperscalers (cloud and AI\/ML platforms), and application platforms (enterprise software and SaaS). M&A activity is not evenly distributed across these layers. Layers 1 and 2 are largely consolidated globally, with established players like Schneider Electric and ARUP competing on contract terms rather than ownership. Layers 4 and 5, hyperscalers and application platforms, are dominated by a handful of tech giants with little appetite to be acquired. The deal flow is compressing into Layer 3: data centre operators. This is where asset ownership, power rights, and customer relationships converge and where the gap between what a platform is worth inside a conglomerate and what it commands as a standalone entity is widest. Operators in this layer sit at the critical interface between the physical infrastructure being built by Layers 1 and 2, and the demand generated by Layers 4 and 5, making them the natural target for investors seeking exposure to the structural growth story without building from scratch.<\/p>\n

The value chain reveals not just how the industry is structured, but where buyers are focusing and why the three investment principles \u2014 platform exposure, time-to-capacity, and structured pathways \u2014 are becoming critical.<\/p>\n

Platform exposure matters more than assets.<\/strong> In Layer 3 (Operators), the companies attracting the most interest are not those with the largest single facilities, but those with replicable platforms<\/strong> and the ability to scale. ST Telemedia Global Data Centres (STT GDC)<\/strong> exemplifies this: headquartered in Singapore, it has built a platform with 2.3GW of design capacity across 12 markets. Since KKR and Singtel first invested in 2024, its pipeline has grown from 1.4GW to over 1.7GW, demonstrating the kind of platform-level capability buyers seek\u2014management that can navigate permitting across multiple jurisdictions and secure power for successive builds. In Malaysia, Empyrion Digital<\/strong> secured approval for a 200MW hyperscale campus in Johor on a “rare 34.9-acre site,” making it one of the few projects to gain both site approval and power allocation despite tightened government approvals. These are not one-off assets; they are development engines.<\/p>\n

Time-to-capacity is the new currency.<\/strong> As supply constraints tighten, delivery track record commands a premium. In Layer 2 (Contractors), Gamuda’s ability to complete the AIMS data centre in just eight months\u2014versus competitor bids of 12-16 months\u2014illustrates why execution speed is now a valuation differentiator. For buyers, acquiring a platform with an embedded delivery capability means acquiring the ability to bring capacity online faster than building from scratch. This is the core logic of “buy mode”.<\/p>\n

Structured pathways are emerging as the default entry mechanism.<\/strong> Across the region, minority stakes and JVs are giving investors a foothold with a clear path to control later. The KKR and Singtel acquisition of STT GDC is the definitive example: the consortium first invested S$1.75B through preference shares and warrants in 2024, then returned in 2026 to acquire the remaining 82% stake at a S$13.8B enterprise value. This structured entry allowed KKR and Singtel to secure exposure to the platform while deferring full control until the asset had demonstrated further growth. In Malaysia, Digital Realty’s acquisition of CSF Advisers (owner of TelcoHub 1 in Cyberjaya) represents another pathway: entering through an existing operational asset with immediate connectivity value\u2014TelcoHub 1 hosts over 40 network providers and connects to AWS, Google, and major internet exchanges\u2014while acquiring adjacent land for up to 14MW of future expansion.<\/p>\n

The tailwind: Construction-phase value precedes operational-phase M&A.<\/strong>
\nToday’s capital expenditure by hyperscalers is flowing to contractors like Gamuda and to platforms securing approved sites like Empyrion Digital. But this construction phase is creating the assets that will become tomorrow’s M&A targets. For investors, this means two things: first, platform-level exposure now means identifying operators like STT GDC that are building scalable pipelines; second, structured pathways like the KKR\/Singtel playbook allow investors to participate in the construction phase while positioning for control once assets are operational and stabilized.<\/p>\n

III. Close-Up: Why Singapore and Malaysia Anchor the Region\u2014And Where the Next Wave Forms<\/span><\/h2>\n

As M&A activity intensifies, capital is concentrating most densely in two markets that form the region’s established core: Singapore, the mature premium hub, and Malaysia, its fast-scaling neighbour. Understanding what makes these two markets attractive is essential for underwriting any deal\u2014but the real opportunity lies in recognising that the tailwinds powering them are now spreading across the rest of Southeast Asia.
\nThe relationship between Singapore and Malaysia runs deeper than geography. Formally established in January 2025, the Johor-Singapore Special Economic Zone (JS-SEZ) creates a cross-border operating environment that specifically advantages data centre operators in Johor serving Singapore-based customers. The zone offers a special corporate tax rate of 5% for up to 15 years for qualifying activities, explicitly including AI and quantum computing supply chains, compared to Malaysia’s standard corporate tax rate of 24%. The IMFC-J’s (Invest Malaysia Facilitation Centre \u2013 Johor) single-window permitting system, operational since February 2025, has streamlined approvals across federal and state agencies, and the RTS Link (Rapid Transit System) due by end-2026 will ease cross-border workforce mobility. The practical effect for data centre operators is compelling: Johor assets can serve Singapore’s demand at Malaysian cost structures, pairing Singapore-grade connectivity with land and power that the city-state itself cannot provide. The JS-SEZ, in effect, transforms the Johor-Singapore corridor into a single investable geography.<\/p>\n

According to Bloomberg’s score for data centre investment attractiveness\u2014which weights factors like power, connectivity, market size, and cybersecurity\u2014Singapore and Malaysia are the clear leaders in Southeast Asia.\"Charts<\/a>Power availability is the lifeline of any data centre, and here the gap is most pronounced. The 2022 Smart Grid Index (SGI), which measures grid reliability, smart technology adoption, and energy security, ranks Singapore first in ASEAN with a score of 75%, and Malaysia second at 71.4%. The next closest ASEAN country trails significantly. This means operators in Singapore and Malaysia face lower risk of outages, greater grid stability, and a clearer path to integrating renewable energy\u2014all critical for the 24\/7 operations and sustainability targets of hyperscalers.<\/p>\n

Data centres exist to move data, and connectivity is their circulatory system. Here, Singapore is unrivalled. Of the 97 international submarine cables currently in ASEAN, 29% land in Singapore. The city-state boasts 27 cable landings, making it the most connected nation in the region. Furthermore, an additional 36 cable landings are planned across ASEAN through 2027, with Singapore again set to receive the newest landings. This dense web of connectivity ensures low latency and high redundancy, which is why global cloud providers place their regional cores in Singapore.<\/p>\n

Malaysia benefits directly from its proximity to this connectivity hub. Terrestrial fibre connections, provided by incumbents like Telekom Malaysia, link Johor’s data centre parks directly to Singapore’s submarine cable infrastructure. This gives operators in Johor the best of both worlds: Singapore-grade international connectivity at Malaysian operating costs.<\/p>\n

The Ripple Effect: Tailwinds Across Emerging ASEAN Markets<\/strong>
\nThe same forces driving Singapore and Malaysia, AI demand, cloud adoption, and the need for data residency, are now creating powerful tailwinds across the rest of Southeast Asia. These markets are at earlier stages of development, which means greater greenfield opportunity and, eventually, consolidation potential.<\/p>\n