- Bank of England Details Unbundling Framework for FundedRe Deals
The BoE's proposed solution involves separating FundedRe into two distinct components for valuation purposes: an investment asset and a longevity swap. - PG&E Talks Up Wildfire Reforms While Analysts Push Buybacks
PG&E outlined expectations for California's new wildfire legislation while fielding investor questions about potential share buybacks amid the utility's depressed stock price. - Insurers Push For Federal Foreign-Backed Litigation Reform Measure
Companies like Allstate and Odyssey Re are backing legislation that would target what some estimate is a $23B problem. - NFIP Becomes Political Talking Point as Shutdown Looms
As Congress battles over funding, the National Flood Insurance Program becomes the latest casualties of political brinkmanship. - Cyclone Pool Stabilizes Suncorp's Reinsurance Costs, But Future Losses Are "In the Hands of the Gods"
Suncorp executives underscored how Australia’s cyclone reinsurance pool is cushioning insurers from catastrophe volatility, even as affordability pressures and climate-driven hazard trends keep risk firmly in focus.
As catastrophe models grow more complex—and climate change accelerates risk—the Risky Science Podcast sat down with Dr. Paul Wilson, Head of Catastrophe and Climate Research at Twelve Securis, to explore how practitioners use, challenge, and refine these tools in insurance-linked securities (ILS) investing.
Wilson, who has spent nearly 20 years in catastrophe modeling, from RMS to Securis, stressed that models are indispensable for providing a “common language” for risk transfer, but never the final word.
From Models to Investment Process
At Twelve Securis, model outputs are always weighed against other information, historical claims, cedent quality, capacity, and contract structure.
Wilson emphasized the importance of a multi-model, multi-vendor approach, not to blend results, but to compare and understand divergences:
- Transactions – Stress-testing key assumptions, probing sensitivities, and identifying red flags in exposure data, whether incomplete or over-specified.
- Portfolios – Building internal adjustments to vendor catalogs, layering climate signals, and refining risk views peril by peril.
Climate-Conditioned Catalogs and Forward-Looking Risk
Wilson is a strong advocate for using climate-conditioned stochastic catalogs to stress test models under specific climate states, such as a warm Gulf of Mexico during El Niño.
This approach, he explained, provides probabilistic “what if” scenarios that give investors a clearer sense of portfolio sensitivity under evolving climate conditions.
Secondary perils like severe convective storms and floods, once considered minor, are now central to portfolio risk. While vendor models are maturing, Wilson cautioned that aggregate contracts tied to frequency perils require precise terms and clear investor appetite.
AI, Exposure Data, and Model Transparency
Artificial intelligence is beginning to reshape exposure data collection and model development cycles. Yet Wilson warned that over-specified inputs, often bulk-filled by AI systems, can be as damaging as incomplete data.
He also highlighted the need for transparent vendor updates and bespoke internal adjustments: “We rarely make blanket changes. It’s about interrogating assumptions and making detailed, event-level refinements.”
Opportunities Ahead
Looking forward, Wilson pointed to upcoming major upgrades in U.S. hurricane and earthquake models as a pivotal moment for the industry. Managers who have invested in their own internal risk frameworks will be best positioned to navigate the changes and maintain investor confidence.
🎙️ Get the full episode of the podcast on your favorite player here.
