Climate and Nature Financial Risk
The Global Risks Alliance (GRA)
Climate and nature financial risk is measured with models that leave much out: tipping points, compound and cascading events, nature, social and political responses, and the feedback between finance and the economy. Gaps can lead institutions and supervisors to understate risk. This page sets out where the main gaps are, how they arise, what research is closing them, and who decides how to treat what is unknown.
GRA · In development · service level R0
Climate financial model gaps are the parts of climate and nature risk that current models omit or simplify, such as tipping points, compound events, nature and feedback between finance and the economy. Gaps can lead to understated risk. Supervisors, risk functions, research funders and boards decide how to treat them.
How satellite radar turns a flood into an exposure update. Synthetic scene; no real event or place.
The working screen behind this subject: its structure, terms and states. Names are placeholders and values illustrative.
Model and measurement gaps · Nature-sensitive sectors · LEAP cycle 2
| Agriculture | 3 | High | High | High | High |
| Food processing | 1 | High | Medium | Medium | Medium |
| Mining | 1 | High | Low | Medium | Medium |
| Real estate | 15 | Medium | Low | Low | Low |
Geolocation, uncertainty and aggregation. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Abrupt changes such as ice sheet loss or forest dieback are rarely in economic damage functions, so severe outcomes are underweighted.
Simultaneous or sequential hazards, such as drought then fire, and their effects on supply chains are harder to model than single perils.
Few models capture how financial losses reduce lending, which slows the economy and raises losses further.
Point estimates from complex chains of models look precise. Without uncertainty ranges, users treat them as more reliable than they are.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Tipping point | A threshold beyond which a system changes abruptly and often irreversibly. |
| Compound event | A combination of hazards occurring together or in sequence with amplified impact. |
| False precision | Presenting an uncertain estimate with more precision than the evidence supports. |
Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.
| Indicator | Unit | Cadence | Source type | Label |
|---|---|---|---|---|
| Scenario exercises including tipping points | count | annual | Supervisor and research publications | REPORTED |
| Disagreement between physical risk providers | correlation of scores | per study | Comparison studies | ESTIMATED |
| Models with second-round effects | count in supervisory exercises | annual | Exercise documentation | REPORTED |
| Nature risk metrics in use | count by framework | annual | Framework reviews | REPORTED |
| Uncertainty ranges reported with results | share of published results | per review | Disclosure reviews | OBSERVED |
| Research outputs on climate-financial modelling | count | annual | Bibliographic databases | REPORTED |
The topic as a computational graph: named sources feed the methods, an evidence review gates the outputs, and the decisions stay with the authorities that hold them.
Fig. 02Nodes carry their type, two metric rows and a state; connectors show what feeds what; dashed frames name the owners; the timeline is the cadence. Design of the pipeline, not a record of a run.
Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.
| Source | What it gives | Licence class |
|---|---|---|
| NGFS technical documentation and research | Scenario limitations and research agenda | Free, NGFS terms |
| IPCC Sixth Assessment Report chapters on tipping points and compound events | Assessed evidence on low-likelihood high-impact outcomes | Free, IPCC terms |
| Institute and Faculty of Actuaries reports on climate scenarios | Critiques of scenario damage estimates | Free to read, terms to check |
| FSB and BCBS reports on climate data and methodologies | Gaps in data and methods | Free, FSB and BIS terms |
| Academic comparisons of physical risk scores | Divergence among providers | Free to read, terms per publisher |
Climate and Nature Financial Risk prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| Who decides | The question | Evidence this page serves |
|---|---|---|
| Supervisors and the NGFS | How to incorporate tail risks and second-round effects | Research findings and exercise experience |
| Risk functions | How much judgement overlay to apply to model results | Gap analysis and uncertainty evidence |
| Research funders | Which modelling gaps to prioritise | Gap assessments and user needs |
| Boards | How to use results whose uncertainty is large | Ranges, scenario spreads and limitations |
Physical risk depends on precise location: a few metres of elevation can determine whether a building floods. Yet financial data often record only an address, a postcode or a company headquarters rather than the operating assets. Hazard data vary in resolution and may not reflect future climate. Damage functions relating hazard to loss are uncertain and often derived from limited data.
Errors compound as models are chained: climate projections feed hazard models, which feed damage models, which feed financial models. Aggregating results into a single score can conceal both the uncertainty and the concentration that matter most.
Linking assets to precise locations, using spatial statistics and geostatistics and Earth observation. Coverage remains incomplete.
Reporting ranges and sensitivity rather than single values. Communicating uncertainty clearly is difficult.
Comparing models with observed losses and with each other, using scenario design and stress testing. Validation against the future is impossible.
Checking whether aggregation hides concentration or tail risk.
The current service level is R0: information and enquiries. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.