Fault and Flood
Response

AI Fallback Method Aims to Keep Cyclone Insurance Payouts Flowing

Liberty Mutual Re and Celest.Science are developing an AI-powered fallback methodology for parametric tropical cyclone insurance to ensure claims can be settled when primary weather data is unavailable.

Liberty Mutual Re and Celest.Science are developing an AI-powered fallback methodology for parametric tropical cyclone...

Liberty Mutual Reinsurance has partnered with climate intelligence firm Celest.Science to develop a fallback methodology for parametric insurance when key weather data is unavailable. The collaboration aims to ensure continuity in claims settlement for tropical cyclones, a growing concern as climate change intensifies these storms. Liberty Mutual Reinsurance, known as LM Re, is part of the Liberty Mutual Insurance Group. Its partner, Celest.Science, specializes in artificial intelligence, climate physics, and advanced modelling.

Their joint effort will explore combining LM Re's parametric risk transfer expertise with Celest.Science's climate modelling capabilities. The focus is on creating a reliable alternative for assessing tropical cyclone events. This is crucial for situations where traditional observational data from weather stations or satellites is delayed, disrupted, or simply missing.

How the fallback approach works

Parametric insurance triggers a payout based on the occurrence of a pre-agreed event, like a storm reaching a specific intensity at a defined location, not on a traditional assessment of physical damage. The speed and transparency of these payouts depend entirely on access to trusted, continuous weather data to verify the trigger.

Celest.Science applies AI and climate physics to fill data gaps where direct observations are limited. The proposed fallback methodology would use multiple independent data streams alongside advanced physical and statistical modelling to characterise tropical cyclones. Using several sources aims to reduce dependence on any single one. This supports greater confidence, objectivity, and operational continuity in determining if a parametric insurance trigger has been met.

Why data continuity matters for parametric products

For parametric insurance to function, the data verifying the event must be indisputable and available. Any disruption to that data flow can delay or even prevent payouts, undermining the product's core promise of swift financial recovery after a disaster. LM Re is actively exploring alternative approaches to ensure this data continuity when standard inputs are disrupted.

Jean-Christophe Garaix, Head of Parametrics & Agriculture at LM Re, explained that clients need timely access to funds for recovery when tropical cyclones strike. Léo Lemordant, Co-Founder and CEO of Celest.Science, noted that extreme weather often occurs precisely in regions where observational networks are fragmented or incomplete. While parametric catastrophe bonds typically specify a backup data provider in their contracts, many parametric insurance agreements do not. Identifying and formalizing additional data sources to act as fallback providers is therefore seen as a valuable step for the entire parametric re/insurance market.

Next steps and validation requirements

The collaboration between LM Re and Celest.Science is now entering a phase of technical assessment. Any future use of this methodology in LM Re's insurance products is not guaranteed. It first requires successful technical validation to prove its accuracy and reliability. Following that, the approach must pass internal governance reviews and have specific contract terms finalized. The partners will assess the methodology's potential to strengthen the resilience of parametric insurance and support its goal of fast, transparent claims settlement. The entire initiative hinges on securing these validations and approvals before integration into any product offering.

Topics

#Response

Related coverage

More from Response