This interactive tool demonstrates how the Confidence Approach can support underwriting decisions during the North Atlantic Hurricane Season, where forecasts are uncertain and decisions carry significant financial risk.

The methodology applies principles from Brian Hill and Richard Bradley's Confidence Approach. It combines seasonal hurricane forecasts with a decision-theoretic framework that considers both the consequences of a decision (the stakes) and an organisation's attitude towards uncertainty (its level of caution). Rather than seeking a single "correct" forecast, the approach evaluates how strongly the available evidence supports different decisions.

The tool is designed to help organisations explore how different attitudes towards uncertainty influence underwriting and portfolio decisions. It allows users to compare alternative decision strategies using the same forecast information, making the impact of different assumptions clear and ensuring that the rationale behind each decision is transparent and explicit.

Developed in collaboration with the London School of Economics and Political Science (LSE), this project was originally funded by the Lighthill Risk Network. The application incorporates seasonal hurricane forecasts collated by the Barcelona Supercomputing Center and historical hurricane data from the International Best Track Archive for Climate Stewardship (IBTrACS)

Visit the app here.

Select your level of caution (risk appetite), define the financial stakes of your decision and explore how different seasonal hurricane forecasts or historical data influence the recommended underwriting strategy.

Technical features of the Decision Theoretic Tool
1. Coverage
North Atlantic Hurricane Season.

2. Historical data
International Best Track Archive for Climate Stewardship (IBTrACS) tropical cyclone archive (1900–2025).

3. Forecasts
Seasonal hurricane predictions from different forecasters (universities, government agencies, private companies and other organizations) collated by the Barcelona Supercomputing Center.

4. Decision methodology
The Confidence Approach combines:
a. uncertainty in the forecasts
b. the financial stakes of the decision
c. the user's attitude towards uncertainty 
to identify which decisions are most strongly supported by the available evidence. 

Resources to continue learning
1. Article: Why decision theory? by Tom Philp. Read here.
2. Article: Environmental decision making under uncertainty by Roman Frigg. Read here.
3. Academic paper: Making confident decisions with model ensembles. Roussos, Bradley & Frigg (2021). Access here. 

Visit the app here.