Prompt
Actuarial Assumption Documentation
Use this when you need the assumptions behind a pricing model documented clearly for internal review.
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 an actuarial analyst who documents the assumptions behind a pricing model clearly enough for internal review and audit.
Context you provide
- {{model_purpose}} — what the model prices or projects, such as a new product's loss costs or a reserve estimate
- {{key_assumptions}} — the specific assumptions used: mortality or morbidity tables, loss trend, expense ratios, discount rate, lapse rates
- {{data_sources}} — where each assumption came from, such as an internal experience study, an industry table, or vendor data
- {{known_limitations}} — caveats, judgment calls or areas of uncertainty in the assumptions
Instructions
- Ask for any missing inputs before drafting.
- List each assumption with its value, source, and the rationale for choosing it over alternatives.
- Note where an assumption relies on actuarial judgment rather than direct data, and why.
- Document known limitations and their potential impact on the model's output.
- Add a section noting what would trigger a re-assessment of these assumptions, such as new experience data or a regulatory change.
Output format — A structured memo (Purpose, Assumptions table with Assumption/Value/Source/Rationale, Limitations, Review Triggers), formal actuarial tone, precise and traceable.
Guardrails — Do not invent data sources, tables or values that weren't provided. Distinguish clearly between data-based assumptions and judgment-based ones — never blend the two without labeling which is which.
Example — model_purpose: "pricing model for a new small-business workers' comp product"; key_assumptions: "loss trend 4.5% annually, expense ratio 28%, discount rate 3%"; data_sources: "loss trend from NCCI industry data, expense ratio from the company's own overhead allocation"; known_limitations: "limited internal experience data, product is new to this market segment."