What is a reporting agent in a parametric insurance contract?
A parametric contract pays on an index, not a loss adjuster's estimate. Wind speed crosses a threshold at a named station, or rainfall over a basin clears a trigger level, and the payout moves. That sounds clean until you ask the question every cedant and every reinsurer eventually asks: who says the index actually crossed the line?
That's the reporting agent's job.
The role, stripped down
A reporting agent is the party named in the contract to source, verify, and report the data the trigger depends on. In a wind cover, that might mean pulling the official best-track file from a weather agency and confirming the storm's path and intensity against the named stations in the contract. In a crop or flood cover, it might mean compiling rainfall totals or river gauge readings from the stated source and certifying them against the attachment point.
The reporting agent doesn't set the trigger and doesn't decide the payout. The contract already fixes both. What the agent does is confirm, in writing, that the data says the event happened, or didn't. That report is what the calculation agent uses to run the math and what both sides use to agree the claim without a dispute process eating the settlement timeline, which is the whole reason a buyer picked a parametric structure in the first place.
Some contracts fold the reporting and calculation functions into one party. Others split them: a reporting agent sources the data, a separate calculation agent applies the payout formula. The split matters less than the next point.
Why the market cares so much about independence
A reporting agent appointed by the cedant looks like the cedant grading its own homework. One appointed by the reinsurer looks the other way. Market practice has settled on third-party agents precisely because neither side wants to defend a number their own appointee produced, especially on a large loss where the payout runs into real money and the official bulletin everyone would prefer to cite hasn't been published yet.
This is also where the single-source problem shows up. If the contract's only data source is one ground station and that station goes offline mid-event, which happens more often than anyone likes during the exact conditions a trigger is designed to catch, there's no fallback read. A reporting agent whose job is limited to transcribing one feed can't solve that. What settles the question fastest is an independent observation of the peril footprint itself: something that doesn't depend on the station staying powered or the official report being timely.
Satellite-based verification closes that gap by giving the reporting agent, or the calculation agent downstream, a second, independent line on what the ground looked like before and after the event.
Where an independent observation fits
Say a named-peril wind cover triggers, but the loss adjustment is contested because the primary station's wind record has a gap during the storm's peak hour. The reporting agent still needs something to point to. A wide-swath multispectral read of the affected footprint, timed to the event and compared before and after, gives a third-party observation that doesn't rely on that one station having stayed online.
This is the gap FlyPix AI's Trigger Verification product is built around: an independent read of the peril footprint from satellite imagery, triggered the moment a qualifying event occurs, rather than relying on a single ground station or one side's own account of what happened.
A few straight answers
Is a reporting agent the same as a calculation agent? Not always. Some contracts combine the two roles in one appointee; others split sourcing the data from running the payout formula on it. Read the contract's definitions section, not the market convention, to know which applies to your deal.
Does the reporting agent set the trigger level? No. The trigger and attachment point are fixed in the contract language before the agent is ever appointed. The agent reports against a threshold someone else wrote.
Who appoints the reporting agent? It varies by deal, but most structures appoint a party independent of both the cedant and the reinsurer, specifically so neither side is grading its own event.
If your contract's language on the reporting data source feels thin, especially the fallback when the primary feed goes dark, it's worth asking what an independent satellite read would add before the next event, not during it.