Per-occurrence versus aggregate protection, and why loss volatility, not habit, should drive the decision
By Luke Renz, ACI
One of the very first questions to ask when evaluating captive viability should be the desired amount of risk the captive is to retain. If the risk tolerance is too low, the captive may not be the appropriate solution as the cost of reinsurance may essentially place the insured back in the commercial market. If the retention is too high, a single severe claim, or an unusually bad year, can consume surplus the captive spent years building.
Too often, the retention is set without much consideration: it mirrors the deductible program from the prior commercial program, it is the standard within a particular group arrangement, or it is whatever amount the reinsurance market quoted first. A better approach starts with the shape of the organization's losses and the captive's capacity to absorb them. This article walks through the core building blocks of retention design and the questions owners should ask before signing off on an attachment point.
Most captive programs are assembled from two core components. Each answers a different question about how much risk the captive should hold.
Simplified illustration of how retention components fit together. Actual structures vary by line, reinsurer, and program.
Owners sometimes frame the choice as one or the other. In practice, most well-designed programs use both, and the real decision is how to calibrate each one relative to the captive's loss profile and capital.
Protects against: Severity. One claim that is far larger than anything in the loss history.
Most critical for: Lines with long tails and heavy-tailed outcomes, such as auto liability, general liability, and excess layers, where social inflation can push individual claims well beyond historical norms.
Key question: What is the largest single loss the captive could pay without impairing its surplus or its ability to pay other claims?
Protects against: Frequency. A year in which ordinary claims arrive in greater numbers than expected.
Most critical for: High-frequency lines such as workers' compensation, auto physical damage, and property deductible programs, where no single claim is threatening but the total can swing meaningfully from year to year.
Key question: If retained losses come in well above expectations, how much of that overrun can the captive absorb before it needs to be capped?
The two protections also interact. A low per-occurrence retention trims the largest claims out of the retained layer, which reduces year-to-year swings and can make aggregate protection less necessary, or cheaper. A high per-occurrence retention leaves more large-loss volatility inside the captive, which makes a well-placed aggregate cap more valuable.
Read the Fine Print: Retentions and aggregates depend on definitions. Confirm whether allocated loss adjustment expense (ALAE) erodes the retention or is shared pro rata above it, how "occurrence" is defined for multi-claimant events, and whether the aggregate attachment is a fixed amount or a percentage of premium or payroll that adjusts at audit. Two quotes with the same headline retention can behave very differently in a loss.
The right attachment point is not a universal number. It is a function of how predictable the organization's losses are and how much variability the captive can carry. Several factors should shape the decision:
| Retention Approach | Reinsurance Cost | Retained Volatility | Surplus & Collateral Needed | Often a Fit For |
|---|---|---|---|---|
| Lower attachment | Higher | Lower | Lower | Newer captives, thin capital, severity-driven lines, limited loss data |
| Moderate attachment with aggregate cap | Moderate | Moderate, capped | Moderate | Established captives with credible loss history and mixed frequency/severity exposure |
| Higher attachment | Lower | Higher | Higher | Well-capitalized captives with stable, high-frequency losses and a long track record |
Directional comparison only. The appropriate structure depends on actuarial analysis of the specific program.
Review losses capped at several candidate retention levels and identify the large losses that drive results. This shows how much of historical cost sits in the working layer versus the occasional large claim.
Ask the actuary to project expected retained losses and the range of outcomes at each candidate retention, including results at higher confidence levels, not just the expected case.
For each option, compare an adverse-year retained loss to the captive's surplus. The structure should leave the captive solvent and compliant with domicile capital requirements even in a poor year.
Compare the reinsurance premium saved by moving the retention up against the additional expected losses and volatility the captive would take on. Factor in collateral and fronting costs alongside reinsurance pricing.
A retention that suited a first-year captive may be too conservative once surplus has grown, and one that suited a soft reinsurance market may be expensive in a hard one. Treat the attachment point as a decision to re-examine each year, not a fixed feature of the program.
There is no single correct attachment point. The right retention is the one that keeps predictable losses inside the captive, transfers the volatility the captive cannot comfortably absorb, and does both at a reasonable total cost. Per-occurrence retentions manage severity, and aggregate protection manages frequency. Getting the balance right requires credible loss data, sound actuarial analysis, and a willingness to revisit the structure as the captive matures.
C.I. provides AM Best-rated fronting paper for captive programs and works alongside your broker, captive manager, and all other service providers to structure retentions and reinsurance that fit your loss profile and capital.
Reach out to C.I. today for a no-cost evaluation of your program.