S&P flags hyperscale data centres as key risk pools

Rising site values expose capacity limits, aggregation concerns

S&P flags hyperscale data centres as key risk pools

Transformation

By Roxanne Libatique

Hyperscale data centres are emerging as a major concentration of insurable risk, changing how insurers and reinsurers deploy capacity and structure covers for technology-related infrastructure in Asia and worldwide, according to S&P Global Ratings. The rating agency estimates that annual investment in data centres could exceed US$300 billion by 2027. Individual hyperscale campuses may carry total insurable values of about US$20 billion to US$30 billion per location when construction and operational phases are combined. Those values are pushing carriers to reassess aggregation, co‑insurance structures, and the treatment of technology assets across multiple lines of business.

Hyperscale sites compared with traditional infrastructure

S&P compares hyperscale data centres with large infrastructure works such as bridges or tunnels, where insurance limits often fall in the US$5 billion to US$10 billion range. By contrast, some of the largest data centre projects are expected to reach US$10 billion to US$30 billion in insurable construction values alone. Each campus typically involves several categories of stakeholders, including large technology firms, specialist developers and contractors, power and utility providers, equity sponsors, and a mix of public and private lenders.

These parties can have differing insurance requirements, spanning construction all-risks, delay in start‑up, property, liability, and cyber insurance. For commercial and specialty carriers, this is contributing to higher exposures and premium volumes across construction, property, casualty, and technology lines. S&P estimates there are about 11,000 operating data centres globally, with a combined insurable asset base of more than US$2 trillion. The number and scale of facilities are expected to rise as new campuses are developed to support cloud computing and artificial intelligence, including in Asian markets such as Singapore, Hong Kong, Tokyo, and key Indian and Southeast Asian hubs.

Premium growth and business interruption exposure

According to S&P, rising demand for data centre insurance could generate around US$10 billion in new premiums in 2026, roughly twice the size of the global aviation insurance market in annual premium terms. The exposure, however, extends well beyond core property values. IT and power infrastructure can represent a substantial portion of total site values. On the operational side, a significant share of risk relates to business interruption driven by outages, dependency on power supply, and operational disruption. For hyperscale operators, S&P notes that business interruption cover is more complex to structure than for conventional property owners because downtime is tied to computing capacity, energy use, and the interconnection of multiple sites. Given these features, S&P expects large technology groups to retain a material share of their exposure through captive insurers. Commercial insurers and reinsurers are likely to participate through layered programs and syndicated placements rather than single‑carrier solutions.

Capacity deployment and collaborative structures

S&P reports that some large commercial insurers now offer per‑risk limits in the low single‑digit billions of dollars for data centre exposures, reflecting the scale of assets on individual campuses. Even at these levels, no single carrier is able to assume the full exposure at the largest sites. To align capacity with project needs, the market is making greater use of multi‑carrier structures led by an insurer or broker. In these arrangements, insurers and reinsurers share risk through layered and quota share reinsurance. Such structures are being used to set consistent terms and conditions across stakeholders and to address concentration and counterparty requirements from lenders and investors.

Over time, S&P expects alternative capital to participate as the risk profile becomes more defined, potentially via insurance‑linked securities or other capital markets mechanisms. Across the project lifecycle, cover typically includes construction and delay in start‑up, operational property damage and business interruption, project cargo for equipment movement, and casualty covers such as third‑party liability, workers’ compensation, professional indemnity, and environmental liability. Technology and cyber policies are expected to adjust as new operational and cyber incident scenarios emerge.

Aggregation risk and limits to insurability

S&P expects insurers to grow their data centre portfolios cautiously, given limited long‑term loss data and the evolving nature of the risk. Compared with traditional infrastructure, hyperscale projects introduce additional complexity, including fast construction schedules, high concentrations of technology equipment, and multiple aggregation channels. Aggregation risk can arise from supply chain disruption, natural catastrophes, and cyber events. Campus-style layouts and geographically concentrated footprints can add further accumulation, particularly in areas exposed to natural perils or grid constraints. According to S&P, carriers with the necessary technical capabilities, modelling tools, and capital resources are likely to underwrite a larger share of this business, echoing patterns seen in cyber insurance.

Despite growth in premium volumes, S&P does not expect full risk transfer to the insurance market. Capacity constraints are likely to limit coverage for hyperscale projects as total insurable values per site reach US$20 billion to US$30 billion. The agency expects that significant portions of operational‑phase exposure – especially business interruption and technology-related losses involving IT equipment – will remain self‑insured or only partially insured, prompting greater use of captives and, potentially, alternative capital structures.

Wider implications for capital, power systems, and ratings

S&P links the insurability of digital infrastructure to broader questions of capital formation, financing costs, and project feasibility, as data centres support AI, cloud services, and enterprise operations. The agency projects that data centres will account for about 14% of US power demand by 2030, up from 5% in 2025, implying increased investment needs across power generation, networks, and related infrastructure. Similar issues are emerging in Asia, where several jurisdictions are balancing data centre expansion with grid capacity, energy mix, and environmental constraints.

For the insurance sector, S&P indicates that global re/insurers started 2026 with solid capital positions and that immediate rating effects from data centre exposure are unlikely, as retained lines per carrier remain limited. At the same time, the agency points to rising limits and underlying exposures as areas for ongoing monitoring. It highlights underwriting expertise, policy wording (particularly for business interruption), reserving practices, claims response, and reinsurance structures as factors that will influence how these risks are reflected in ratings. S&P notes that how re/insurers manage concentration and aggregation risk arising from large clusters of assets and interconnected stakeholders will be an important element in future credit rating analysis, including for groups building or expanding data centre books in Asia.

Related Stories

Keep up with the latest news and events

Join our mailing list, it’s free!