Self-Funding Health Benefits at 100–500 Employees: A Practical Primer

Self-funding means the employer pays its employees’ actual medical claims instead of paying an insurer a fixed premium. A TPA typically runs the plan day to day, and stop-loss insurance caps the worst-case scenario. The idea is old and unglamorous; what makes it interesting is arithmetic that starts working in an employer’s favour somewhere around one hundred covered lives.

The arithmetic of a risk pool

Health claims are volatile for individuals and predictable for groups. One person’s annual medical cost can swing from zero to something enormous; the average across several hundred people moves far less. Statisticians call this the law of large numbers, and it is the entire basis of insurance pricing. A company with 20 employees cannot predict next year’s claims with any confidence. A company with 300 usually can, within a reasonable band.

A fully-insured premium is priced above expected claims — it must fund the insurer’s administration, capital requirements, and margin. An employer whose own claims experience is stable is, in effect, paying a volatility fee for volatility it no longer has. Self-funding is a decision to stop paying that fee and to keep the difference in years when claims come in at or below expectations.

Stop-loss: the piece that makes it survivable

No sensible mid-sized employer self-funds without stop-loss cover. It comes in two forms, usually bought together:

• Specific stop-loss reimburses the plan when any one person’s claims exceed a chosen threshold — protection against a single catastrophic case.
• Aggregate stop-loss reimburses the plan when total claims for the year exceed an agreed ceiling, typically set at 110–125% of expected claims — protection against a generally bad year.

With both in place, the employer’s true exposure is the corridor between expected claims and the aggregate ceiling — a known, budgetable number rather than an open-ended risk.

What the employer gains beyond price

• Cash flow: money leaves when claims are actually paid, not in advance; reserves stay with the company until needed.
• Data ownership: the plan’s claims data belongs to the plan, enabling informed decisions about design, networks and wellness spending.
• Plan design freedom: benefits can reflect the actual workforce — its demographics, locations and utilisation — rather than a standard product.
• No premium tax on the self-funded portion in many jurisdictions, and no funding of an insurer’s margin in good years.

The honest list of risks

Self-funding is not a free lunch, and it fits badly in some situations. Claims will exceed budget in some years even with stop-loss; a company without the balance-sheet strength to absorb a bad quarter should not take that on. Administration quality now matters directly — a weak TPA means slow payments and unhappy employees, and the employer owns that outcome. Fiduciary and compliance responsibilities sit with the plan sponsor. And small headcounts, volatile workforces, or a single site with correlated health risks all weaken the statistical foundation the model rests on.

A reasonable rule of thumb: self-funding starts to be worth analysing at around 100 employees, becomes genuinely attractive for many companies between 150 and 500, and deserves a proper feasibility study — several years of claims data, quotes for stop-loss, and a candid look at cash reserves — before anyone signs anything.