Новая система прогнозов устранит базовый риск в параметрическом страховании от ураганов

A new study co-authored by catastrophe modelling and climate analytics specialist Reask, has demonstrated how advanced downstream forecast analytics can eliminate basis risk in parametric insurance by funding pre-landfall protective actions before a hurricane strikes. Published in the Bulletin of the American Meteorological Society (BAMS) by researchers from Reask, the US Naval Research Laboratory, and NOAA, the study introduces LiveCyc, a probabilistic forecasting system that takes any agency forecast and works out the wind distribution at the 1-km scale of a single neighborhood. To validate the model’s reliability for underwriting automated payouts, the study reconstructed what LiveCyc would have produced ahead of 16 major U.S. landfalls that took place between 2017 and 2024, including Hurricanes Laura, Michael, Ian, Ida, and Milton. Using the real-time NHC forecasts issued before each event, and validated against 1,840 surface wind gust observations, Reask highlighted how LiveCyc managed to demonstrate exceptional accuracy. At the Day-2 decision window, (30 to 48 hours before landfall), Reask explained that 87% of observations fell within the modeled 90% confidence interval. As well as this, the system also achieved a median bias of 0.5 kt and a Continuous Ranked Probability Score of 7.5 kt.. Thomas Loridan, Chief Science Officer at Reask and lead author of the study, commented: “People making decisions about specific sites need a number for each of them, and some sense of how much to trust it. A probabilistic forecast at neighborhood scale gives them both. We put the back-tests in an app so anyone can check the calibration themselves rather than take our word for it.” Importantly, as it’s clearly showcased that LiveCyc can reliably generate 1,000 realisations at 1-km resolution, the study establishes the mathematical foundation required to structure automated pre-landfall parametric insurance triggers. “Because the output is a distribution rather than a single number, the trigger for protective action can be tuned to what that action costs. The paper works this through on two illustrative scenarios drawn from published naval base figures, following the US Navy’s Tropical Cyclone Conditions of Readiness framework. Evacuating non-essential personnel pays off at an 8% chance of exceeding 60 kt, while confirming a ship sortie waits until that chance reaches 33%. The same optimization applies wherever the cost of acting can be quantified against the cost of the loss it prevents,” Reask explained. David Schmid, Global Head of Data Products, Reask, commented: “To embed this into a holistic risk management approach, that very same threshold can be used to trigger a pre-landfall payout from a parametric insurance cover, sized to match the cost of the protective action. That’s the beauty of this structure: because the payout is tied to the same forecast threshold that triggers the action, basis risk is essentially designed out. Once the cover fires, the protective measures are set in motion and the costs are incurred — whether or not the storm ultimately turns away, the payout matches the cost of those measures.” He continued: “To make the cover more comprehensive and also address the scenario when the storm does make landfall, we recommend pairing this with a second post-landfall parametric trigger based on the realized event. The limit is split between the two according to the insured’s cost profile, typically weighted toward the post-landfall trigger, since recovery and reconstruction costs usually exceed the cost of protective action. The forecast leg funds anticipatory action; the realized leg funds recovery.” Reask also works with a number of leading insurance-linked securities (ILS) managers, providing them with its modelling and analytics capabilities. View all of our Artemis Live video interviews and subscribe to our podcast. All of our Artemis Live insurance-linked securities (ILS), catastrophe bonds and reinsurance video content and video interviews can be accessed online. Our Artemis Live podcast can be subscribed to using the typical podcast services providers, including Apple, Google, Spotify and more.