The capital expenditure (capex) super cycle is a major global investment cycle in which capital is shifting from asset-light software and services towards large physical infrastructure. Investment is flowing into energy systems, electricity grids, renewable energy, AI data centres, semiconductor fabrication and advanced manufacturing.

This shift is creating larger, more capital intensive and more interconnected assets. Global energy investment alone is expected to reach USD 3.4 trillion in 2026, while the five largest US hyperscalers – large technology companies that operate extensive cloud and data centre infrastructure – are expected to invest more than USD 800 billion in AI related capital expenditure in 2026. For insurers, the capex super cycle is increasing the amount of insurable value while also changing the nature of risk, with exposures becoming larger, more concentrated and more difficult to diversify. 

AI data centres and renewable energy investment could generate around USD 200 billion in cumulative commercial property and casualty insurance premiums between 2026 and 2030.

AI data centres could generate around USD 91 billion in cumulative premiums, while renewable energy could generate around USD 111 billion.  These premiums represent almost entirely new business created by the current investment cycle, rather than incremental exposure growth from inflation or GDP. The opportunity extends across construction and operation, with the larger and more durable opportunity coming from operational risks as new infrastructure creates property, business interruption and liability exposures.

Risk accumulation occurs when multiple insured exposures can be affected by the same event or disruption. In the capex super cycle, the sigma 3/2026 report identifies four major channels through which risk can accumulate: single risk severity, geographic concentration, supply chain bottlenecks and network dependencies. 

For example, hyperscale data centres can cluster in locations with access to energy while depending on common suppliers, electricity, telecommunications and cloud infrastructure. A single disruption can therefore affect multiple insureds, sectors and lines of business at the same time. 

For insurers, this can make losses more correlated and severe, particularly with business interruption as a driver. Managing these risks requires detailed mapping of supplier networks, infrastructure dependencies and logistics bottlenecks.

Renewable energy investment is creating substantial new insurance demand across both the construction and operational phases of projects. Swiss Re Institute estimates that renewable energy investment could generate USD 111 billion in new cumulative insurance premiums between 2026 and 2030, around 22% higher than the estimated premiums from AI data centres. 

By 2030, renewable energy could generate around USD 26 billion in annual premiums, with USD 23 billion coming from operational assets and USD 3 billion from construction. The operational opportunity spans solar, wind and hydropower, with Europe representing the largest regional opportunity, just ahead of Asia. 

The biggest insurability challenges are risk quantification, diversification, maximum loss and the economic viability of coverage. AI data centres and other large infrastructure projects can concentrate billions of dollars of value in individual sites while depending on shared infrastructure, networks and complex supply chains. 

AI data centres illustrate the challenge particularly clearly. Individual campus values of many AI data centres now exceed USD 10 billion and can reach as much as USD 50 billion in replacement value. Once operational, business interruption, contingent business interruption and liability become increasingly important alongside physical property risk.

Insurability, rather than the availability of capital, is the main constraint because insurers need to be able to understand, quantify and manage increasingly large, concentrated and interconnected risks before they can deploy capacity with confidence. 

Unlike previous waves of industrial development, the capex super cycle can strain multiple dimensions of insurability at once. Larger and more interconnected assets increase maximum loss potential and reduce diversification, while limited operating histories for emerging technologies make losses harder to estimate. 

These challenges do not mean the risks are uninsurable. Large commercial risks such as AI data centres and energy infrastructure remain fundamentally insurable through private market solutions. Advances in engineering expertise, underwriting, risk modelling and programme design can be used to unlock capacity.