Published October 2026.
Direct Answer: Aggregate stop-loss reinsurance pays when a cedent’s loss ratio on a defined subject book rises above an attachment point (for example 70%) and reimburses losses within a corridor up to a limit (for example 90%), so recovery equals subject premium multiplied by the ratio band actually exceeded, capped at the corridor width. On $100M of subject premium with 70% attachment and 90% limit, the treaty pays nothing at a 65% loss ratio, pays $10M at 80%, and pays a capped $20M at 105% because the 20-point corridor is fully consumed. Stop-loss is earnings protection: it smooths the combined ratio and shields underwriting profit in bad frequency years, unlike per-occurrence excess-of-loss, which is built to absorb large individual losses against surplus.
On a treaty slip, stop-loss looks like other aggregate covers until you see attachment stated as a percentage of premium rather than a dollar threshold on a single loss. That distinction drives recovery modeling and program design. For treaty context, see quota share, excess-of-loss, and aggregate structures, then use the sections below to decode loss-ratio mechanics on the slip in front of you.
Stop-loss in one paragraph
Aggregate stop-loss (also stop-loss ratio reinsurance) covers the whole subject portfolio over a defined period, usually one accident or underwriting year. The cedent sums incurred losses under the treaty definition, divides by subject premium (the slip’s premium base), and obtains a loss ratio. When that ratio crosses the attachment point, the reinsurer pays within the band up to the limit (the corridor). Losses below attachment and losses above the limit stay with the cedent unless another layer applies. Subject premium may differ from statutory written or earned premium; every ratio point converts to dollars through that denominator, so confirm the slip’s premium definition before pricing. For why cedents transfer risk at all, see how carriers transfer risk to reinsurers and policyholders.
The loss-ratio corridor — how attachment and limit work
Read a stop-loss slip as a corridor between attachment and limit in loss-ratio points. At 70% attachment and 90% limit, the corridor is 20 ratio points (90% minus 70%). Each point equals 1% of subject premium within the payable band.
Why the corridor is in ratio points, not dollars
Per-occurrence excess-of-loss attaches at dollar retentions on individual claims. Stop-loss attaches on book experience relative to premium scale: the same dollar attachment would mean different protection for a $40M book and a $140M book. Ratios keep cover proportional to the subject portfolio.
Corridor width in dollars = Subject premium × (Limit − Attachment), with attachment and limit as decimals (0.70 and 0.90).
On $100M subject premium and a 70% to 90% corridor:
- Attachment dollars = $100M × 0.70 = $70M incurred before the treaty pays within the band.
- Limit dollars = $100M × 0.90 = $90M incurred at which additional corridor recovery stops (subject to slip definitions).
- Maximum recovery = $100M × (0.90 − 0.70) = $20M.
Recovery (standard corridor, no coinsurance or reinstatement) = Subject premium × [min(Actual loss ratio, Limit) − Attachment] when Actual loss ratio > Attachment; otherwise zero. At 80% actual loss ratio: 80% − 70% = 10 points; recovery = $100M × 0.10 = $10M. The cedent still funds the first $70M and any loss above $90M on worse years.
Worked example on a $100M book
Annual aggregate stop-loss: subject premium $100M, attachment 70%, limit 90%, ultimate incurred on the subject book. Three scenarios:
Scenario A — 65% loss ratio (below attachment)
Incurred losses = 65% × $100M = $65M. Ratio is below 70% attachment. Treaty recovery = $0. Cedent retains all $65M.
Scenario B — 80% loss ratio (inside the corridor)
Incurred losses = 80% × $100M = $80M. Points above attachment = 80% − 70% = 10. Recovery = $100M × 0.10 = $10M. Losses from $70M to $80M ($10M) are reimbursed. Net loss after recovery = $80M − $10M = $70M (attachment dollars).
Scenario C — 105% loss ratio (above the limit)
Incurred losses = 105% × $100M = $105M. Uncapped points above attachment = 35; limit caps the payable band at 90% − 70% = 20 points. Recovery = $100M × 0.20 = $20M. Loss above 90% limit = $105M − $90M = $15M retained. Cedent net after recovery = $105M − $20M = $85M.
| Actual loss ratio | Incurred losses | Ratio above attachment | Treaty recovery | Cedent net losses (after SL) |
|---|---|---|---|---|
| 65% | $65M | 0 (below 70%) | $0 | $65M |
| 80% | $80M | 10 points | $10M | $70M |
| 105% | $105M | 20 points (capped at limit) | $20M | $85M |
Stop-loss buys defined ratio bandwidth, not unlimited tail. Program placement with quota share and working XoL is covered in the reinsurance treaty complete guide.
Why it is called earnings protection, not capital protection
Stop-loss responds to aggregate underwriting results, not a single shock loss against surplus. Earnings protection means shielding the underwriting P&L and combined ratio when frequency or moderate severity pushes the loss ratio into the corridor. A cedent may remain adequately capitalized while missing earnings targets; stop-loss dampens that income-statement volatility on the subject book.
Capital protection is the usual frame for per-occurrence excess-of-loss and catastrophe covers that absorb losses large enough to stress surplus, models, or regulatory minimums. A $250M nat-cat event is severity and accumulation territory; a 70/90 stop-loss does not replace cat XoL. See catastrophe portfolio management, accumulation, and PML reinsurance for that stack. Quota share shares every loss dollar from the first dollar; stop-loss sits above a ratio retention and pays only when experience deteriorates into the corridor—cedents buy it when they accept “normal” bad years to attachment but want reinsurance for incremental ratio deterioration that would distort reported margin.
Stop-loss vs aggregate excess-of-loss
Both can respond to aggregated experience; the trigger differs decisively.
Aggregate excess-of-loss attaches when aggregate dollars of loss exceed a dollar attachment (often after per-occurrence terms). Example wording: pay 100% of aggregate losses excess of $50M, subject to an aggregate limit of $25M. The trigger is cumulative incurred amount, not loss divided by premium.
Aggregate stop-loss attaches when the loss ratio exceeds a percentage of subject premium. The same $50M of losses may trigger aggregate XoL on a small book but sit below 70% attachment ($70M) on a $100M premium book.
Stop-loss scales with premium automatically; aggregate XoL holds a fixed dollar retention unless indexed. Cedents thinking in margin and ratio targets (MGAs, niche carriers) favor stop-loss; dollar-stated retentions (facilities, captives) favor aggregate XoL. Treaty taxonomy appears in the reinsurance complete professional guide.
When cedents buy it in the underwriting cycle
Stop-loss demand rises when earnings visibility matters and ratio capacity tightens—not usually as the first defense in a soft market with cheap working layers.
Hard markets. Reinsurers shrink lines and raise attachments on proportional and working XoL; cedents may retain more frequency and buy stop-loss to cap the ratio tail. Fewer markets on 70/90 classes pushes rate-on-line relative to the $20M corridor cap in the $100M example. Cycle context: insurance underwriting cycles, hard and soft markets, coverage, and pricing.
MGAs and thin margins. Programs with explicit loss-ratio targets to carriers align attachment to the tier where profit share or commission erodes, protecting economics in a heavy frequency year without repricing every underlying policy.
Results season. Carriers bind before year-end to limit fourth-quarter development surprises on volatile lines. Ratio attachment grows with subject premium mid-year if the book expands—a deliberate feature.
January renewal. Stop-loss often renews with the rest of the program; collateral and retro pressure affect whether prior 70/90 economics hold. See the global reinsurance market, Lloyd’s, Bermuda, and January renewal. Reconcile premium forecast, expected loss ratio, and corridor cost: a $20M limit on $100M premium adds little if modeled ratios never reach 70%; it matters when stress cases cluster at 85% to 95%.
FAQ
What is the difference between stop-loss and excess-of-loss reinsurance in one paragraph?
Excess-of-loss reinsurance pays when an individual loss or a defined occurrence exceeds a dollar retention, or when an aggregate dollar total exceeds a threshold, depending on the layer type. Aggregate stop-loss reinsurance pays when the incurred loss ratio on a subject book exceeds a percentage attachment point and reimburses the cedent within a ratio corridor up to a limit. Excess-of-loss is built to cap severity or aggregate dollar exposure; stop-loss is built to cap underwriting margin deterioration relative to premium scale.
How does the loss-ratio corridor work on a stop-loss treaty?
The attachment point and limit are expressed as loss ratios on subject premium. The corridor is the difference between limit and attachment in ratio points. Each ratio point equals 1% of subject premium in dollar terms. Recovery equals subject premium multiplied by the actual loss ratio above attachment, capped so the ratio credit does not exceed the limit. On $100M subject premium with 70% attachment and 90% limit, the corridor is 20 points and maximum recovery is $20M.
Show the worked numbers: how does a stop-loss respond on a $100M book at 65%, 80%, and 105% loss ratios?
Subject premium $100M, attachment 70%, limit 90%. At 65% the incurred loss is $65M, below attachment, recovery $0. At 80% incurred loss is $80M, which is 10 ratio points above attachment, recovery $100M × 0.10 = $10M. At 105% incurred loss is $105M; uncapped points above attachment would be 35, but the limit caps payable band at 20 points from 70% to 90%, recovery $100M × 0.20 = $20M, with $15M of loss above the 90% limit retained by the cedent.
Why is stop-loss called earnings protection rather than capital protection?
Stop-loss smooths the combined ratio and protects underwriting profit in bad frequency or moderate severity years when the loss ratio enters the corridor. It does not replace catastrophe or large per-occurrence excess-of-loss designed to absorb shock losses that threaten surplus and regulatory capital. A cedent can remain adequately capitalized while still suffering an earnings miss; stop-loss targets that earnings volatility on a defined book rather than solvency stress from single large events.
What is the difference between stop-loss and aggregate excess-of-loss?
Aggregate stop-loss attaches when the loss ratio exceeds a percentage of subject premium and pays within a ratio band up to a limit. Aggregate excess-of-loss attaches when cumulative incurred losses exceed a dollar amount and pays subject to aggregate limits and any per-occurrence terms in the slip. Stop-loss scales automatically with premium; aggregate XoL retains a fixed dollar attachment unless the treaty is indexed or amended.
When in the underwriting cycle do cedents buy stop-loss cover?
Cedents most often buy or renew stop-loss when hard markets tighten capacity, when MGAs and program carriers need to protect thin margins against frequency-driven ratio spikes, and when management wants earnings stability into results reporting. Demand rises when cedents retain more on working layers but still want a defined ratio tail cover on a subject book, commonly aligned to January or other annual treaty renewals.