Appearance
Triggers & payouts
This page explains how an area's drought index is converted into a payout amount. It feeds Pricing the risk.
The two thresholds
Each area's payout curve is defined by two thresholds, derived as percentiles of its own history:
- Trigger: where payouts begin. A season better than the trigger pays nothing. The reference product uses the 10th percentile.
- Exit: where payouts reach their maximum (the full sum insured). The reference product uses the 0th percentile (the worst observed season).
Between trigger and exit, the payout scales with how severe the season was.
Severity
For a given season's score, severity is how far it has fallen from the trigger toward the exit, capped at 0-1:
if score is better than the trigger:
severity = 0 → no payout
else:
severity = (trigger − score) / (trigger − exit) capped at 1A severity of 0 means "just reached the trigger"; 1 means "at or beyond the exit, full severity."
Payout structures
Severity does not translate straight into a payout. A payout structure with a threshold (the reference product uses 0.30) shapes the final payout. The two structures behave very differently:
Franchise: payout = severity if severity ≥ threshold, else 0
Deductible: payout = severity − threshold (never below 0)| Franchise | Deductible | |
|---|---|---|
| The threshold acts as | A gate | A subtraction |
| Below the threshold | No payout | No payout |
| At/above the threshold | Pays the full severity | Pays severity minus the threshold |
| Maximum payout (at exit) | 100% of sum insured | 100% − threshold, e.g. 70% |
TIP
Same data, very different price. In the reference project, switching only the payout structure from franchise to deductible drops the portfolio pure premium from 16.16% to 10.95%: because the deductible permanently shaves the threshold off every payout.
Worked example
A season produces severity 0.5 for an area, with a 0.30 threshold:
- Franchise: 0.5 ≥ 0.30, so payout = 0.50 → 50% of sum insured.
- Deductible: 0.5 − 0.30 = 0.20 → 20% of sum insured.
If severity were 0.25 (below the threshold), both structures pay 0.
The payout curve, visually
payout
rate
1.0 | ___________ exit (max)
| /
| franchise step -> /
thr |·············______···/
| | /
0 |_____________|______/________________ drought score
better trigger worse exitWhen you price a product, the platform draws this per area: the score line, the dashed trigger and exit lines, and dots marking the historical years that would have paid out. See Price a product.
Why this matters for pricing
Replaying every historical season through this curve produces a spread of payouts for each area. The average of that spread is the area's expected loss, the foundation of the premium. That is the subject of the next page.
Where to go next
- Pricing the risk, from simulated payouts to the client's price.
- Price a product, adjusting triggers, exits, and structure in the app.
- Glossary, trigger, exit, severity, payout structure.