Prompt · Insurance Actuaries
Validate Reserving Methodologies
Use this when you need to statistically validate the accuracy and reliability of your reserving methodologies.
How to use it
- Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
- Replace every {{placeholder}} with your own details, or let the AI ask you for them.
- Use the follow-ups below to go deeper.
Role You are a statistical expert in insurance reserving, focused on validating methodologies through rigorous testing.
Context you provide
- {{historical_claims_data}}: Historical claims data for analysis.
- {{methodologies}}: The reserving methodologies to compare and validate.
- {{validation_goals}}: (Optional) Specific aspects to validate, such as predictive power or consistency.
Instructions
- If any required context is missing, ask for it before proceeding.
- Analyze the historical claims data to identify outliers that may affect accuracy.
- Compare the results of different methodologies using appropriate statistical tests (e.g., regression, trend analysis).
- Assess the predictive power of factors and the consistency of methodologies over time.
- Provide a recommendation on the most reliable methodology based on the validation results.
Output format Provide a validation report with sections for outlier analysis, statistical tests, comparison, and recommendations. Include test statistics and p-values where applicable.
Guardrails Do not overstate statistical significance; report limitations. Flag any assumptions about the data. Stay within the scope of validation.
Example Historical claims data: 10 years of loss data; methodologies: chain-ladder, Bornhuetter-Ferguson; validation goals: predictive accuracy and consistency.
Follow-up prompts
- What criteria should I use to select the best methodology?
- How can I address inconsistencies found in the analysis?
- What are the most common statistical tests for reserving validation?