Solution: Express the value of a nature-based intervention in the metric insurers use to make decisions - average annual losses - so that community-scale and nature-based place resilience measures can be quantified and recognised in catastrophe models.

Key challenge: Catastrophe models are the lens through which insurers view a place, but while most models capture large-scale defences such as river dykes and coastal flood-wall systems, very few capture community-scale interventions or nature-based solutions effectively - so investment in the resilience of a place cannot be fully recognized in the price of insurance.

Context: The 2018 Camp Fire killed 86 people and destroyed 95% of the structures in Paradise, California, including 11,000 homes, causing $16.5 billion in total losses and $12.5 billion in insured losses, the costliest disaster globally that year. Insurers responded by withdrawing, with non-renewals in California rising 31% in 2019 alone. The withdrawal that followed was not just a pricing failure, it was a signal that the market lacked the tools to assess which properties could be made resilient and which could not. The setting is suburban and semi-rural residential areas in the wildland-urban interface.

Solution in practice: The Nature Conservancy, working with Guy Carpenter, a reinsurance broker, gathered granular data about Paradise: which houses were destroyed, property boundaries from public records, and how much homes cost to rebuild. They fed this into the insurance pricing model, then tested what happened if they added green spaces and parks around the town's edge designed to slow fire spread. The results: these green buffers alone reduced the modelled losses by 27%; when combined with modern building codes, which independently reduce losses by a further 20%, the two measures together produced a compounded reduction of around 42%, cutting total modelled losses from over $10 billion to around $6 billion. That shift would materially improve the risk profile and support more viable insurance outcomes, subject to pricing, market capacity and regulatory conditions. The analysis demonstrates that incorporating place-based resilience measures into insurers’ models can quantify their risk reduction benefits and establish a quantifiable return on resilience investment.

Impact / outcome: Green buffers cut modelled wildfire losses by 27%; combined with modern building codes, about 42% (from over $10bn to around $6bn). Next steps The capability to reflect place-based resilience in modelling therefore exists; what limits its use is not technical capacity but the consistency of demand and the maturity of the evidence base across different contexts, pointing to a role for market innovation alongside data vendors as well as public policy. For example, Colorado's HB 1182 regulation, passed in 2025, requires insurers to account for property, community and landscape-scale mitigation in both pricing and coverage decisions, and offers a model other jurisdictions could adopt to incentivize this behaviour. Location also matters within the intervention itself: the green buffers were 5 to 27% effective depending on where they were placed, so benefit-cost analysis must guide which areas to prioritise. The playbook identifies two areas where innovation is required.

Methodologies fit for the decision being made: place-based risk assessment methodologies can be improved across the property value chain, including robust place-based risk assessment for asset managers with insurability as one indicator of underlying risk, and resilience-adjusted credit analysis for banks, requiring close collaboration among these actors. Clarity on what models can and cannot be used for: climate and catastrophe models serve portfolio-level hazard screening, financial impact assessment across return periods, and pricing or underwriting decisions, but are not designed for site-level or place-based resilience decisions, for which engineering analysis is required - using CAT and climate models for site-level resilience produces confident but incorrect answers.

Organizations involved authoring: Marsh Actors: The Nature Conservancy; Guy Carpenter Implementation partners Data Providers; Insurers; Civil Society Organizations; Local Governments