Probable Maximum Loss (PML)

What Is the Probable Maximum Loss?

The Probable Maximum Loss (PML) is the maximum loss expected with a given probability — equivalently, the loss at a specified return period on the EP curve. The most commonly referenced PML metrics in the (re)insurance industry are the 1-in-100 year, 1-in-200 year, and 1-in-250 year return period losses.

PML vs. Maximum Possible Loss

The PML is sometimes confused with the Maximum Possible Loss (MPL). The MPL is the theoretical worst-case scenario — the loss if every insured property in the affected area were totally destroyed. The PML is a probabilistic concept: it is the loss that would be exceeded only once in the specified return period under the model's assumptions. The MPL is rarely used operationally because it is almost always much larger than any plausible realistic scenario.

Uses of PML in the Industry

• Reinsurance purchasing: An insurer typically buys reinsurance to protect against losses above a retention level, up to its PML. The reinsurance limit is set to cover the difference between the retention and the PML at the chosen return period.
• Capital adequacy: Rating agencies (S&P, AM Best, Moody's) and regulators use PML metrics to assess whether an insurer is holding sufficient capital. An insurer whose 1-in-250 year PML exceeds available capital is considered under-capitalised.
• Accumulation management: Reinsurers use PML to manage their own accumulations — ensuring that no single event can produce a loss exceeding their risk appetite at a given return period.
• Catastrophe bond sizing: The trigger and limit of a catastrophe bond are typically set relative to a PML level — for example, a bond might attach at the 1-in-100 year loss and exhaust at the 1-in-250 year loss.

Spatial Correlation and PML

PML is heavily influenced by the spatial correlation of losses in a portfolio. A portfolio concentrated in a single metropolitan area faces a very different PML profile than the same total insured value spread across many regions. A single earthquake or hurricane can devastate a geographically concentrated portfolio — producing a PML that represents a very high percentage of TIV — while a geographically diversified portfolio would see much lower PML-to-TIV ratios.

Realistic Disaster Scenarios (RDS)

A complement to the probabilistic PML is the Realistic Disaster Scenario (RDS) — a specific, named event used as a stress test. The Lloyd's of London market, for example, requires syndicates to estimate their loss from a set of standard RDS events including a major Florida hurricane, a California earthquake, and a European windstorm. RDS analysis complements probabilistic PML by providing concrete, intuitive scenarios rather than abstract probability levels.