Explore all resources, research, and content related to this topic.
ClimateSET is not a public tool yet. When it is released, the question it is built to answer is whether an asset sits inside a coarse climate-hazard screen for finance — not what the loss would be, and not the insurance terms.
Climate transition risk metrics from different providers can look wildly different for the same firm — but Bingler, Colesanti Senni, and Monnin's peer-reviewed research finds a more nuanced picture: metrics converge notably more for firms that are clearly most or least exposed, and cluster by underlying methodology.
CarbonPlan asked nine climate risk analytics companies for the same fire and flood risk data on 342 real addresses. Seven declined or didn't respond. The two that participated agreed only weakly on which properties were most at risk — and CarbonPlan argues that industry non-participation is itself the finding.
Public climate data is abundant but coarse. Asset-level risk analysis needs resolution public data can't provide, so a private, well-funded market has emerged to close that gap with proprietary, often AI-driven methods. We think that market's structure — who can afford fine-grained data, and who can't — is itself a climate justice problem, made worse when the data isn't even reliably accurate.
GARP tested 13 physical climate risk vendors against the same 100 properties worldwide. The correlation between vendors' flood damage-ratio estimates ranged from 0.2 to 0.9, one vendor geocoded a Boston property 1,507 km away in Atlanta, and dispersion behaves in opposite directions depending on which statistic you use to measure it.
UNEP FI, backed by 44 banks and investors, ran an identical dummy portfolio past 15 climate risk vendors. The same portfolio scored 76.9, 63, and 53.4 out of 100 for physical risk from three different vendors — and physical Value-at-Risk estimates ranged from 0.6% to over 21%.
MSCI's website hosts a PDF of Verdantix's Green Quadrant report on climate financial data and analytics providers — but the version MSCI distributes contains only MSCI's own vendor profile, not the other nine providers assessed. Useful for its evaluation criteria, not for independent comparison.
Seven climate risk vendors assessed the same dummy portfolio and the same flood-prone stretch of road outside Paris — and disagreed almost completely. A detailed Investor Leadership Network study, alongside a 13-vendor GARP benchmark, traces that disagreement to specific, nameable methodological choices, not just irreducible uncertainty.
