Group Captive vs Level Funded for 50 to 150 Employees

Most comparisons of group captives and level-funded plans tell you what each one is. This one compares one health insurance plan funding choice against another and shows what each one costs you in a bad stretch, when financial risk becomes the deciding issue. For a company with 50 to 150 employees, that is the number that decides whether you are still in the plan three years from now.

A level-funded plan wins more often than people expect. A captive loses less when things go wrong. Which one fits depends on your CFO's risk tolerance and which of those two facts matters more.

What both arrangements have in common

Both are forms of self-funding, specifically self insured plans and self funded plans. Both are often used as a transition from fully insured models to self-funding, and level funding gives small to mid-sized businesses a way to self-fund without taking full risk exposure. You pay your own claims up to a limit, where the employer pays claims directly until stop-loss attaches, and you get money back when claims come in under budget. Both are priced on your group's own health risk, not on the community pool that sets small-group fully insured rates.

They are now common. In the KFF 2025 Employer Health Benefits Survey, 37% of covered workers at firms with 10 to 199 workers were in a level-funded plan. At those same firms, 27% of covered workers were in plans KFF classifies as self-funded (KFF, 2025).

The structural differences are short:

• Level funded. One insurance carrier packages admin, a claims fund, stop-loss, and administrative fees into fixed monthly payments and monthly premiums. You are the carrier's customer, and the plans act like fully insured coverage on the surface because the carrier serves as the backstop through stop-loss. If claims run under budget, the carrier returns part of the surplus.
• Group captive. You self-fund in a group medical captive with your own third party administrator and stop-loss, and a middle layer of claims is shared because group captive insurance pools risk among multiple employers in the captive. The Third-Party Administrator manages claims and plan administration. You are a part owner of that pool. If the pool runs well, part of the pooled premium comes back as a dividend.

Both options are alternatives to traditional fully insured plans and fully insured plans priced on the community pool.

The numbers from our 2027 funding simulator

Our health plan funding simulator plays each funding path forward five years, 2,000 times per state, using each state's 2027 small-group rate filings. Seventeen states use final approved 2027 rates and the rest use requested rates. The results are published by state on pages like the Texas 2027 funding analysis by employer size.

Here is the 70-employee result, with the median across the 48 states where both paths are available. "Odds" means the share of simulated runs in which the path cost less than staying fully insured over five years.

Source: Benefitra 2027 funding analysis, conservative pricing basis, generated September 25, 2026.

Read the bad-luck row first

Read the bottom three rows before the top two. A plan that wins nine times out of ten can still put a $167,000 hole in a five-year budget. That hole is what gets a finance team to drop self-funding and go back to fully insured at the worst possible moment.

Part of the gap is an input, and part of it is not

Be careful with the median savings row. In our insurance structure, the median row still reflects quote differences: we price a level-funded package at a premium over the captive for the same group, because that is what we see in real quotes. At 51 to 100 lives the load is 4.5%. So level funded starting behind the captive on the median is partly built in.

The bad-luck row is different. No input sets it directly. It comes out of thousands of simulated claim years, including the 17 named catastrophic conditions the engine draws for each member. The simulator methodology lists every assumption.

If you think our 4.5% load is too high for your market, run your own quote through the simulator. The median gap will shrink, but the bad-year gap will not go away because the funding model behaves differently under stress.

Why the captive's bad year is milder

Many employers look to level-funded insurance or a captive to manage healthcare costs as healthcare costs keep rising and to buffer against rising costs. Most articles point to stop loss insurance when they explain captive safety, because it protects employers from catastrophic claims in both level-funded and captive arrangements. In our model, stop-loss is not where the difference comes from. The level-funded plan actually has the tighter worst-year cap: we model its aggregate corridor at 120% of expected claims, against 125% for the captive.

What happens at the next renewal

The difference is what happens at the next renewal, and the renewal terms are one of the key differences between the two approaches. These are the terms we model:

A level-funded carrier re-underwrites you on your own experience. A year with one cancer case and two NICU stays can push next year's bundled rate up by close to the full size of that bad year, because the company pays fixed costs through that structure even after a bad year. If actual claims exceed what was funded, the carrier re-underwrites based on actual claims.

A captive spreads part of your bad year across the other employers in the pool and caps the increase by contract. In favorable years, surplus funds can also be returned through the pool. You give up some upside in a good year in exchange. That trade shows up in the 5th percentile, not in the median.

If you are already level funded

If you are already level funded and coming off a heavy claims year, our guide to level-funded renewals after a high-claims year walks through what the carrier will ask for.

How the answer changes between 30 and 150 employees

Size changes the trade. The table below covers three sizes. The 30-employee figures use our documented client results setting, while 70 and 150 use the conservative setting, so compare within a row, not across rows. In practice, level funded health plans are more common at smaller sizes, especially for small businesses, while the captive becomes more competitive as headcount grows. At 150 employees the simulator compares the captive with self funded health insurance, because level-funded packages are the default only up to 100 lives.

The 30-employee decision

At 30 employees, the choice comes down to risk tolerance as much as size. A 30-person firm with healthy cash reserves, steady cash flow, and a young workforce can take the level-funded path for the higher typical return; it is simpler to enter at this size, but level-funded plans offer more limited customization options for coverage than a captive and may do less to support a healthier workforce over time. A 30-person firm that would have to cut payroll to absorb healthcare expenses over a five-year period should not.

Where the captive does not win

Two states break the 70-employee pattern. California and Vermont define small group as 1 to 100 employees, so a 70-person group there still buys community-rated fully insured coverage. That price does not move with your claims, so every self-funded path swings harder against it.

The captive still loses less, but in those two states neither path has a positive floor. Check the California 2027 results by employer size before assuming the national pattern applies to you.

New York and Rhode Island are not in our comparison at all. The simulator removes self-funded paths where state rules restrict small-group stop-loss, rather than showing an option you could not buy.

What stop-loss costs in 2026 and why it matters for 2027

Both arrangements use stop loss coverage to protect employers when claims exceed expected thresholds, and that market is hardening. Segal found that medical stop-loss premiums rose an average of 12.7% across 225 plans in its 2026 data, up from 9.7% the year before. It also reported that the number of claimants with seven-figure claims has grown an average of 25% per year over the last four years (Segal, June 2026).

The 2026 Aegis Risk Medical Stop-Loss Premium Survey, reported by IFEBP, put specific stop-loss at a $100,000 deductible at $260.63 per employee per month, with a 13.6% increase at that level. For level-funded plans, that matters because monthly premiums still have to support stop-loss pricing even when claims are stable. The survey's authors said the data "does not indicate any easing on the 2027 renewal cycle" (IFEBP, September 2026).

Medical trend for 2027

Medical costs are rising too. Marsh expects pressure on healthcare expenses and overall plan year pricing to keep total health benefit cost per employee climbing 8.2% on average in 2027, and 11% for employers that make no changes (Marsh, September 2026).

For a level-funded group, those changes are usually folded into bundled fixed monthly payments for the next plan year. A captive proposal usually quotes stop-loss, the captive layer and admin as separate lines, so you can see which one moved. That visibility matters more in a year when stop-loss is rising faster than claims.

Five contract terms to compare before you choose

The table that decides this is not in the proposal summary. Ask for these terms in writing from both bidders, because they shape plan design and determine how much more control an employer really has:

1. Renewal cap. Is there a contractual maximum increase, and does it apply to the stop-loss, the captive layer, or both?
2. Lasers. Can the stop-loss carrier set a higher deductible for one sick employee at renewal? Our guide to stop-loss laser provisions covers what to push back on.
3. Surplus conditions. Is the level-funded refund of unused funds paid only if you renew? Is the captive dividend paid on your results, the pool's results, or both?
4. Collateral and exit. What does the captive hold back when you join, when is it returned, and how much notice do you have to give to leave?
5. Run-out. Who is responsible if actual claims exceed the funded period and are submitted after the plan ends, and for how many months?

A group that gets clean answers to all five can compare the two offers on cost. A group that cannot is comparing marketing.

How to test this on your own group

Our numbers describe a typical group in each state. Yours will differ by age, industry, claims history, provider networks, and workforce profile, not just averages. Here is how to get your own version:

• Enter your census and current health insurance plan data in the funding simulator. Your census is read in your browser and not uploaded.
• Switch on the captive and the level-funded options and set the bad-year view. Compare the 5th percentile rows, not just the medians.
• Plug in any real quotes you have. The simulator recalculates on your numbers in place of ours, and claims data is the main reason to compare how each option may save money or create potential cost savings.
• Read how group captive plans or the group captive model works on our self-funded captive page and how the level-funded model is packaged.

If your CFO can live with the bad-luck number on the level-funded side, it is a fair choice. If they cannot, that number is the reason to look at a captive before your 2027 renewal.

You can see how other employers made funding decisions in our case studies.

Frequently asked questions

Sources

1. Benefitra. "2027 Funding Analysis by Employer Size" (state pages, simulator output generated September 25, 2026). https://benefitra.com/health-insurance-funding/texas/employer-size/
2. Benefitra. "Health Plan Funding Simulator Methodology." 2026. https://benefitra.com/health-plan-funding-simulator-methodology/
3. KFF. "2025 Employer Health Benefits Survey." October 22, 2025. https://www.kff.org/health-costs/2025-employer-health-benefits-survey/
4. Segal. "Medical Stop-Loss Premiums Increase Nearly 13%." June 8, 2026. https://www.segalco.com/consulting-insights/medical-stop-loss-premiums-increase-nearly-13-percent/
5. International Foundation of Employee Benefit Plans (IFEBP). "Stop-Loss Premium Rates Rose a Further 5%+ in 2026" (reporting the 2026 Aegis Risk Medical Stop-Loss Premium Survey). September 9, 2026. https://blog.ifebp.org/stop-loss-premium-rates-rose-a-further-5-in-2026/
6. Marsh. "Employers expect health benefit costs to jump 8.2% in 2027, and the impact will likely be felt by workers, according to Marsh." September 2, 2026. https://www.mercer.com/en-us/about/newsroom/employers-expect-health-benefit-costs-to-jump-in-2027/