The idea in one paragraph
Captive insurance means forming (or joining) an insurance company that you own, so premiums that would have gone to a commercial carrier's shareholders instead build up in an entity that returns underwriting profit to you. Large corporations have run single-parent captives for decades across property, liability, and benefits lines. What's changed is the middle market: group medical stop-loss captives now let employers with as few as 50–100 covered lives band together, pool a layer of health-claims risk, and share in the results.
The volume of search interest in "captive insurance" reflects that shift — for benefits buyers, the practical question is almost always the group medical captive, so that's what this guide covers.
How a group medical stop-loss captive works
A member employer in a group captive still runs its own self-funded plan: it pays predictable, everyday claims from plan assets, exactly as any self-funded employer does. The layers above that change:
- The employer's retained layer — claims up to its specific deductible, funded by the plan.
- The captive layer — a band of risk (for example, claims between the specific deductible and roughly $250,000–$500,000) that all member employers pool together inside the captive. Premiums for this layer stay in the captive; if the pool's claims come in under expectations, the surplus is returned to members as dividends.
- The reinsurance layer — catastrophic claims above the captive band, passed to a commercial stop-loss or reinsurance carrier.
The structural point: the layer where mid-sized claims live — the layer commercial stop-loss carriers price most conservatively — becomes something members own rather than rent. Good collective claims experience is money back instead of a carrier's margin.
Who captives fit — and the honest trade-offs
Group captives tend to reward employers that are:
- Mid-sized — commonly ~50–500 covered employees, big enough to self-fund but too small to absorb stop-loss volatility alone
- Committed to managing the plan — captive membership usually comes with expectations: data sharing, cost containment programs, engaged brokers
- In it for the medium term — captives smooth results across years and members; a one-year tourist mentality defeats the design
- Comfortable with governance — members typically post collateral and participate in captive oversight
The trade-offs are real: capital is tied up as collateral, joining and exiting take diligence, heterogeneous member quality can drag results, and a badly run captive is just a complicated way to share other people's claims. The vetting question is always the same — how does this captive select and hold members accountable?
Why claims performance matters twice in a captive
In a conventionally insured plan, reducing claims helps at renewal. In a captive it pays twice: lower claims in the employer's retained layer are direct plan savings, and lower claims across the pooled layer become captive underwriting profit — returned to members as dividends.
That double payback changes the economics of claims-reducing benefits. Redirecting avoidable ER visits with a $0-copay telehealth and in-home diagnostics benefit like EZaccessMD trims the retained layer and keeps escalated episodes out of the captive band — the fever evaluated at home tonight instead of the observation stay that would have hit the pooled layer. Captive members and their brokers can model the claims-side impact with the ROI calculator, or zoom out with the cost containment guide.