Parametric insurance pays a pre-agreed amount when a measured index crosses a threshold, regardless of the actual loss. The payout depends on the parameter, not the damage. That removes the claims adjuster and turns settlement into a data lookup.
Mechanics: index, trigger, payout, oracle
A contract names an index (wind speed, ground acceleration, rainfall over a window, snowfall by a date), a trigger threshold, a payout structure, and a data source that settles it. Parametric payments are triggered by a transparent, observable index, while indemnity insurance is triggered by a dollar loss amount assessed by a claims adjuster. Because no investigation is needed, payouts can be made within days of a triggering event. The data source acts as an oracle: an automated weather station, a seismograph network, or a modeled hazard footprint. For weather-index crop cover, payouts are triggered when rainfall measured by automated weather stations deviates from predefined thresholds during key crop growth phases.
Catastrophe pools: CCRIF and ARC
CCRIF SPC, the Caribbean catastrophe risk pool, settles fast because no adjuster estimates damage. Since 2007 it has made 54 payouts to 16 member governments totaling roughly US$244.8 million, all within 14 days of the event. Its largest single payout was nearly US$40 million to Haiti in August 2021 after the 7.2 magnitude earthquake. The African Risk Capacity runs drought cover on a rainfall and crop-water index. ARC paid Senegal US$23.1 million for the 2019 agricultural drought when a rainfall and crop-water index breached its trigger. Mexico moved its sovereign cover into capital markets, with FONDEN issuing catastrophe bonds covering earthquake and hurricane risk.
Business interruption without physical damage
A ski resort needs snow to open. It can buy cover that pays if snowfall falls short of a threshold by a set date, with no building damage involved. The US ski industry and associated businesses lose roughly US$1 billion in low-snow years. A weather-data provider such as NOAA verifies the trigger, and payouts are automatic with no site visits or loss adjusters. The same structure covers rained-out events, lost foot traffic, and other economic losses that indemnity policies, keyed to physical damage, would not pay.
Basis risk
The payout tracks the index, not your loss. Basis risk arises when the index does not correlate with the actual loss. You can suffer real damage and collect nothing if the index does not trip, or collect with no loss at all. A quake just outside the measured radius, or rain that misses the station but not your fields, leaves the policy silent. Index transparency and trust drive uptake: Kenyan smallholders value mechanisms that reduce basis risk, though they value transparency about how rainfall is measured even more. One vendor forecast projects the market near US$29.3 billion by 2030 at about 9.9% CAGR, a market-research estimate rather than a settled figure.
Try it
Design and backtest a snowfall trigger from public NOAA data. Pull a daily snowfall record for one station from NOAA’s Climate Data Online (GHCN-Daily, the SNOW element), say a 20-year history for a station near a ski area. Set a strike: cumulative snowfall below X inches by December 31. Set a payout: a fixed sum if the strike is breached. Backtest how many of the 20 seasons would have paid. Then proxy actual loss with season-total snowfall (or skier-visit data if you can find it) and plot index outcome against the proxy. The seasons where the index pays but the proxy shows a normal year, or the reverse, are your basis risk made visible. Effort: a half day in a spreadsheet or pandas.