Stop loss captives are not one-size-fits-all structures. Different models allocate risk, capital, and underwriting gains in different ways — which directly impacts how (and when) participants receive earnings.
Structure
Multiple unrelated employers participate in a shared captive.
Typical layers:
All participating employers share risk in the captive layer.
How Earnings are Distributed
After the policy year:
Distribution methods:
In some models:
Key Facts
Structure
Each employer participates in its own legally segregated “cell” within a larger captive.
Risk layers:
Each employer’s results are financially separated.
How Earnings are Distributed
Because risk is isolated:
Distribution methods:
Key Facts
Risk and reward are individualized, not pooled.
This model is attractive to employers who want:
Structure
A large employer forms its own captive insurance company.
The captive:
There are no unrelated participants.
How Earnings are Distributed
Because the employer owns the captive:
There is no shared distribution formula.
Key Facts
Complete control and complete ownership of results.
Structure
A consultant, broker, or trade association sponsors the captive.
Employers join under common underwriting guidelines.
Risk may be:
How Earnings are Distributed
Common approaches include:
1. Tiered Distribution
2. Corridor-Adjusted Allocation
3. Experience-Based Weighting
Often:
Structure
The captive takes a fixed percentage of the stop loss layer.
Example:
Risk is partially shared between captive and reinsurer.
How Earnings are Distributed
Because the captive holds proportional risk:
This model can smooth volatility because the captive does not hold 100% of the working layer.
Structure
Employers share a defined risk corridor collectively.
Example:
How Earnings are Distributed
Distribution depends on corridor performance:
This structure emphasizes shared mid-layer performance.
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.