Stop-Loss Reinsurance

Stop-loss reinsurance is non-proportional reinsurance that protects a ceding insurer against its total losses for a period, usually one year, exceeding an agreed level. The level is normally expressed as a loss ratio, so a typical cover might pay the losses that take the insurer's loss ratio above 105% of net earned premium, up to 130%. It is also called excess of loss ratio cover or, more broadly, aggregate excess of loss.

Ordinary excess of loss reinsurance looks at each risk or each event separately. Stop-loss looks at the whole year's result, so it responds whether a bad year comes from one disaster or from thousands of ordinary claims. It is written as treaty reinsurance and should not be confused with medical stop-loss insurance bought by self-funded employer health plans, or with a stop-loss order used in securities trading.

Imagine a family that budgets $6,000 a year for groceries and makes a deal with a relative: if the year's grocery bills go above $6,600, the relative pays the extra, up to $8,000 in total. It doesn't matter whether the overspend came from one huge party or from many small shopping trips; only the year's total counts.

The family still feels a bad year, because the first $600 of overspending is theirs. But it knows that a disastrous year cannot ruin it. Stop-loss reinsurance gives an insurer the same kind of floor under its annual results. Because the protection only begins once the insurer is already losing money, the insurer still has every reason to keep its spending, or its claims, under control.

The insurer's loss ratio for the period is:

The stop-loss recovery, for an attachment point $a$ and an exhaustion point $e$, is:

• $L$ is the insurer's net incurred losses for the period, sometimes including loss adjustment expense.
• $\text{EP}$ is net earned premium for the period.
• $a$ and $e$ are the attachment and exhaustion loss ratios, written as decimals.

Suppose an insurer earns $200 million of premium and buys cover from 105% to 130%. If losses are $240 million, the loss ratio is 120% and the recovery is 15% of $200 million, or $30 million. If losses reach $280 million, the loss ratio is 140%; the recovery is capped at 25% of premium, $50 million, and the insurer bears the remaining $20 million above the cover in addition to the $210 million below it. Covers often carry a monetary cap as well, so that rapid premium growth does not inflate the reinsurer's exposure.

The attachment point is usually set above the insurer's break-even loss ratio, so the insurer has already lost money before the reinsurer pays. This limits moral hazard: a stop-loss attaching too low would let an insurer underprice or loosen its underwriting without bearing the consequences. Many covers also require the insurer to keep a co-participation, for example 10% of every loss inside the layer.

Stop-loss suits lines where results swing with conditions affecting a whole portfolio rather than with single large claims. Crop and hail insurance is the classic example, since one dry or stormy season can raise claims across thousands of farms at once; parametric covers and weather derivatives address the same exposure from a different angle. Small mutual insurers, captives and health insurers also buy stop-loss to steady results and protect their regulatory capital.

In the United States, many employers pay their staff's medical claims themselves instead of buying a fully insured group policy. They protect the plan with stop-loss insurance, which comes in two parts: specific stop-loss, which pays claims above a set amount for any one person, and aggregate stop-loss, which pays when total plan claims exceed a set level, often around 120% to 125% of expected claims. Strictly, this is insurance bought by the employer rather than reinsurance bought by an insurer, but the aggregate cover works on the same principle. State insurance regulators oversee the stop-loss carriers, and some states set minimum attachment points to stop very low attachments being used to sell what is in effect a fully insured plan.

• Cost and scarce capacity. Reinsurers are wary of covering an insurer's whole result, since it bundles pricing, reserving and catastrophe risk together, so cover is expensive and often limited.
• Delayed settlement. The final loss ratio depends on the claims reserve, so the recovery may not be known for years in long-tail lines.
• Definitions matter. What counts as premium, which losses are included and how other reinsurance recoveries are netted off all change the outcome, and disputes often turn on these wordings.
• Not a substitute for discipline. Insurers usually combine stop-loss with quota share and per-risk cover, rather than relying on it to fix poor results or to reduce the year-to-year volatility of a weak portfolio.

• National Association of Insurance Commissioners (NAIC)
• International Association of Insurance Supervisors (IAIS)