Average Severity

Insurers price risk by answering two questions: how often do bad things happen, and how much does each one cost? Average severity answers the second, and paired with claim frequency it is half the arithmetic of every premium ever quoted.

The calculation is disarmingly simple: divide total incurred losses by the number of claims in the period. If a motor insurer pays $10 million across two thousand claims, average severity is $5,000.

The discipline lies in defining claims and losses consistently, because legal costs, partial payments and reopened files all blur the count. Severity data also demands patience before conclusions, since one bad quarter can reflect a single large loss rather than a trend, so analysts smooth severity over rolling years and separate catastrophe-driven spikes from the underlying cost drift before repricing.

Severity trends drive premiums as surely as frequency does. Modern cars cost more to repair because sensors and cameras hide behind every bumper, so motor severity climbs even when crash counts fall, and medical inflation does the same for health and liability lines.

The National Association of Insurance Commissioners tracks frequency and severity together in its auto insurance database reports because the pair explains why premiums differ across states, and regulators watching affordability care whether rising costs come from more accidents or from pricier ones. Insurers manage severity with different tools than frequency.

Repair networks, parts sourcing, medical fee schedules and litigation strategy all attack the cost per claim, while telematics and safety campaigns attack how often claims occur, so knowing which lever matters requires watching the two measures separately. Buyers can act on severity too, because contracting with repair networks before a loss, using approved medical providers and documenting assets thoroughly all shrink the eventual claim cost, and companies that treat severity as manageable, rather than weather, pay for less risk over time.

Severity distributions are famously skewed, so the average alone can mislead. Catastrophe years expose that skew violently, as one storm can produce a handful of claims larger than an entire quiet year's total and drag the average far above the typical claim, which is why actuaries study the distribution, not just the mean, when setting reserves and prices.

Reinsurers price almost entirely off the tail of severity, which is why primary insurers buy cover for the claims the average hides. For business insurance buyers, severity thinking changes programme design: a company with a few large potential losses needs high limits and cares about the tail, while one with many small losses focuses on deductibles and claims handling.

Reserving practice leans heavily on severity estimates, because actuaries project the ultimate cost of open claims by blending observed severity with development patterns, and a sustained rise in severity forces reserve strengthening even if the number of claims is stable. Earnings surprises at insurers often start as severity surprises in one line of business.