Anatomy of the Gap
Basis risk is the difference between what a parametric policy pays and what the buyer actually loses. It is the single most important concept in the field, and the one most often waved away in sales conversations. Master it and you can evaluate any parametric structure on sight. Ignore it and you will eventually buy a policy that pays handsomely for a storm that spared you and nothing for the one that ruined you.
The gap has structure. It arises from three distinct sources, and a serious buyer diagnoses each.
The first is geographic. The index is measured somewhere specific — an airport weather station, a river gauge, a grid cell in a satellite dataset. Your facility is somewhere else. The hazard at the measurement point and the hazard at your building are correlated but not identical. A tornado's damage is notoriously local; winds that flatten one block leave the next intact. Rainfall can differ sharply across a few miles. The farther your facility sits from the reference point, and the more spatially erratic the peril, the wider this component of the gap. This is why a river gauge twenty miles upstream of your site is a weaker basis than one on your own reach of the river, and why wind products anchored to a distant station carry more basis risk than those anchored nearby.
The second is what we might call structural or design basis risk: the mismatch between the index and the mechanism of your loss. Suppose your data center's true vulnerability to a hurricane is storm surge flooding the ground-floor electrical rooms, but your parametric policy triggers on wind speed. A storm can deliver catastrophic surge with wind speeds that never reach your trigger, or howling winds with a surge that stays offshore. You have insured the wrong variable. The index and the loss are driven by related but separable physics. Choosing an index that actually governs your dominant loss pathway — surge height if flooding is your exposure, not wind speed — is the core design decision, and getting it wrong is the most common and most expensive error in the field.
The third is temporal and definitional: the fine print of how the index is measured. Is wind speed defined as a one-minute sustained average or a three-second gust? Measured at what height? Over what averaging window? A policy triggering on sustained wind and a policy triggering on peak gust will pay differently for the same storm, and a facility whose damage comes from gusts is poorly served by a sustained-wind trigger. Degree-day accumulations depend on the base temperature chosen. Flood triggers depend on gauge datum and timing. These definitional choices are where basis risk quietly hides, and where the diligent buyer earns their keep by reading the schedule.
These three sources combine into a total gap that can fall in either direction. When the index under-reports your loss — you are damaged but the trigger is not met — you have negative basis risk, and it is the direction that hurts. You paid premium, you took the loss, and you collect nothing, exposed at the exact moment you believed yourself protected. When the index over-reports — the trigger fires but your loss was small — you have positive basis risk, a windfall. The insurer prices the expected value of positive basis risk into your premium; it is your gain and their cost. But negative basis risk is the one that can leave an operator insolvent while holding a policy that technically performed as written.
The uncomfortable truth is that you cannot eliminate basis risk without eliminating the product. Every dial you turn to close the gap — a nearer station, a more sophisticated multi-variable index, a trigger tuned to your exact loss curve — adds cost, complexity, and often the very loss-verification machinery that parametric cover exists to avoid. Push it far enough and you have reinvented indemnity insurance, slowly and expensively. The objective is not zero basis risk. The objective is basis risk small enough, and understood well enough, that the speed and certainty you bought are worth the residual gap. That is a judgment, not a calculation, and it is the judgment that separates a sophisticated parametric buyer from a naive one.