Exceedance probability is the probability that a variable will be at or above a stated threshold within a defined period, usually a year. It is the complement of the cumulative distribution function: , where is the probability of a value below . Its reciprocal is the return period. An annual exceedance probability of 0.01 corresponds to a 100-year event, although the chance of at least one such event in 30 years is , or about 26 percent.
The measure is central to pricing catastrophe risk. Models produce an exceedance probability curve showing the likelihood of losses above each amount. The probability that losses reach the trigger level of a catastrophe bond largely determines its yield, and flood maps define a special flood hazard area by a one percent annual chance. In finance, value at risk applies the same idea. Estimates for rare events are uncertain, especially with fat tails.