Pioneering the Next Frontier of Risk Transfer: ILS and Digital Infrastructure
The Growing Significance of Digital Infrastructure in Re/Insurance
As the fields of artificial intelligence and digital infrastructure continue their rapid growth within the reinsurance industry, the successful integration of insurance-linked securities (ILS) into this burgeoning sector hinges on pivotal structural advancements. Key among these are the creation of standardized parametric triggers and efficient portfolio aggregation mechanisms, as articulated by Michael Moloney, a prominent figure at Oliver Wyman.
Immense Capital Demand for Digital Advancement
Moloney, who serves as Partner, Global Head of Insurance and Asset Management, and Managing Partner, Actuarial, at Oliver Wyman, recently shed light on how the ILS industry can strategically enter the digital infrastructure domain in a report by Guy Carpenter. He pointed out the monumental capital requirements for this sector, with estimates suggesting a staggering US$5 trillion to US$7 trillion in global capital needed between 2025 and 2030 to fuel data center and AI infrastructure expansion. Leading hyperscale companies alone are projected to invest over US$800 billion by 2027. Individual campuses are routinely valued at US$10 billion to US$30 billion, consuming energy comparable to medium-sized cities and financed through complex capital structures involving equity, secured debt, and project bonds from various institutional investors.
Navigating the Obstacles for ILS Integration
Despite the apparent suitability of this environment for ILS and sidecar structures, Moloney cautions that the inherent risks are exceptionally large. He identifies several potential impediments to the ILS market's seamless entry into this space. The most formidable barrier is risk modellability. The existing ILS market relies heavily on exceedance probability curves derived from decades of physical event data and validated by third-party catastrophe models. In contrast, digital infrastructure risks—such as power supply interruptions, PPA counterparty defaults, construction delays, and grid interconnection failures—lack an equivalent actuarial foundation. Moreover, these risks are dynamic; the correlation between power supply disruptions and data center revenue loss constantly shifts with changes in grid topology, contractual agreements, and operational dependencies. The primary insurance market has yet to accumulate sufficient loss experience or develop the analytical frameworks required for large-scale modeling.
Addressing Portfolio Size and Remuneration Challenges
Moloney further emphasizes that the structural complexity extends to sizeability. While individual campus risks are substantial enough for bond tranching, the portfolio diversification assumption underlying traditional cat bond structures—which relies on geographic and peril independence—does not directly translate to digital infrastructure. Regarding remuneration, Moloney acknowledges that attractive yields could draw ILS investors due to genuine supply-demand tension in the market. However, he reiterates that accurate pricing necessitates robust modellability, which circles back to the initial constraint.
The Path Forward: Standardization and Aggregation
The development of the digital infrastructure market will inevitably require time, given the extensive development and modeling infrastructure needed. Moloney draws a parallel with the cat bond market, which took a decade of primary market development, loss experience accumulation, and modeling infrastructure before securitization became widely viable. He predicts a similar trajectory for digital infrastructure risk, albeit potentially accelerated by intense capital pressure and strong commercial incentives. Moloney identifies three key developments that would significantly hasten ILS entry: the emergence of standardized primary insurance products with clear parametric triggers (especially for power supply interruption and PPA performance), which would generate essential loss data and pricing benchmarks; portfolio aggregation across multiple campuses and counterparties, enabling diversification and making tranching feasible even without geographic independence; and treating power supply agreements as sophisticated long-dated financial instruments with defined credit exposure, aligning them more closely with risks ILS investors already understand.
A New Paradigm for Risk Transfer
Moloney points out that the US$123 billion in alternative reinsurance capital, built over two decades, institutionalized an existing risk class. However, the digital infrastructure risk pool is fundamentally different; it is being created from scratch. Its financing structures treat risk transfer as an essential prerequisite for deployment rather than an optional enhancement. Institutional lenders and bond investors in project finance require a clearly defined, transferred, and contractually managed exposure profile before committing capital. In this market, risk transfer is integral to financeability, not incidental. This demand dynamic for ILS growth in digital infrastructure is unprecedented. Unlike the traditional natural catastrophe ILS market, which capitalized on established, well-modeled risks, digital infrastructure is evolving in reverse. The rapid pace of risk generation far outstrips the absorption capacity of traditional insurance balance sheets, even as institutional capital stands ready. The primary constraint is not investor appetite but the absence of foundational primary market infrastructure—actuarial frameworks, standardized products, and structured triggers—necessary to deploy this capital. The speed of ILS integration will depend less on capital supply and more on how quickly the insurance market develops the product architecture for this new asset class. The foundational work in primary insurance markets is currently laying the groundwork for the next chapter of capital convergence.