Which funding structure fits your company?

It can. The problem is what it costs to make it safe.

Self-funding works because most groups have claims near their expected average. That only holds when you have enough claimants for the average to mean something. Twenty employees is roughly 45 covered members. One premature birth, one transplant or one late-stage cancer can exceed the entire year's expected claims by itself.

Stop-loss exists to cap that. But at 20 lives you need a very low specific attachment point to be protected, and low attachment points are expensive. By the time you've bought enough stop-loss to sleep, your fixed cost approaches what fully insured would have charged, and you've taken on cash-flow timing risk and administrative duties for the privilege.

Then there's lasering, where a stop-loss carrier singles out one person at renewal and gives them a far higher deductible than everyone else. A large group absorbs that. A 20-life group can lose three good years to one laser.

Level funding solves the same problem differently. The ceiling is contractual, the surplus still comes back if claims run low, and you still get your claims data. That's why the small-group answer is usually level funded rather than self-funded.

A shelf-rated plan has its rates published in advance, the way a product on a shelf has a price tag. You look them up. Nobody underwrites you (reviews your group's age or health history to set a price), nobody asks for a census (a list of your employees and their ages), nobody asks medical questions.

One published rate per coverage tier, and the tiers are a short list running from employee only up through family. The 24-year-old graphic designer and the 61-year-old welder pay the same. So does a company in rural Montana and one in downtown Manhattan.

That's the opposite of ACA small group, where your rate is built from each employee's age and where they live. Shelf-rated pricing is set against a large national pool instead of your specific twelve or forty people.

Guaranteed issue means coverage cannot be denied because of health history. Eligibility conditions still apply and not every employer qualifies.

Lasering is when a stop-loss carrier singles out one specific person and gives them a much higher deductible than everyone else on the plan.

Here's how it plays out. Your plan has a $50,000 specific deductible, meaning stop-loss reimburses you for any one person's claims above $50,000. At renewal the carrier reviews your claims and sees an employee with a condition expected to run $400,000 next year. They renew you, but they write that one person at a $350,000 deductible while everyone else stays at $50,000. That employee has been lasered. You're now carrying $350,000 of exposure on someone you thought was covered.

The carrier isn't doing anything improper. They're pricing a known cost instead of a probability. It's a legitimate underwriting tool.

The problem is asymmetry. A 900-life employer absorbs a laser. A 40-life employer can lose several good years to one. And lasers show up at renewal, when there's very little runway to shop.

What to ask for. A no-new-lasers provision says the carrier can't add a laser at renewal that wasn't already there. Pair it with a rate cap (a contractual ceiling on how much your renewal can go up) and you've closed both ways a single claimant can blow up a small plan.

These provisions carry a small premium load, often low single digits as a percentage. For a group under roughly 200 lives that's the cheapest protection in the entire structure. If a proposal doesn't state whether lasers are permitted, ask that before you ask about the rate.

On September 3, 2026, CMS and the Small Business Administration announced that the individual coverage health reimbursement arrangement, the ICHRA, is now officially the CHOICE Arrangement, short for Custom Health Option and Individual Care Expense Arrangement.

One precision point a lot of coverage blurs: this was an agency terminology change, not an act of Congress. Bills to write the CHOICE name into federal statute have been introduced several times and none has become law. The underlying ICHRA regulations still govern, unchanged.

Practically, nothing about how the benefit works changed. Existing plan documents stay valid. Employers already offering an ICHRA don't have to do anything. The rules on employee classes, affordability, notice and substantiation are the same rules.

Expect both names in circulation for a while. Plan documents, carrier forms and IRS guidance still say ICHRA in plenty of places.

Most carriers write level funded at 10 to 25 enrolled employees as a floor, with some going lower in specific states. True self-funding with a separate stop-loss placement typically starts around 50, and group captives commonly want 50 to 100 before they'll take a member.

These thresholds are set by carrier appetite, not statute, so they move. A group that was declined two years ago may be writable today.

Within a single metro area, the negotiated price for the same procedure can differ by several multiples between facilities, and price tracks negotiating leverage rather than quality.

Two things follow. The obvious one is unit cost. The one people miss is downstream: higher-performing surgeons produce fewer complications, fewer readmissions and fewer revision procedures. A revision is a second full episode of cost.

The empowering version matters more than the savings version. Most people choosing a surgeon are choosing on a referral and a parking lot. They've never had access to outcome data on the physicians in their own network. Steerage isn't narrowing their choice. It's the first time anyone handed them the information.

It sounds backwards. The employer pays more per prescription and saves money. Three mechanisms.

Acquisition cost instead of spread. The drugs on a broad $0 list are almost entirely generics. Under a traditional pharmacy contract those are priced off a discount from a benchmark, and the difference between what the plan is billed and what the pharmacy is paid stays in the middle. Sourcing the same molecule at acquisition cost plus a transparent dispensing fee removes that layer.

Abandonment. A significant share of new prescriptions are never picked up, and cost at the counter is a leading reason. A prescription that's never filled produces no drug spend and no clinical benefit. At $0, the fill rate goes up.

The downstream math. Maintenance medications for blood pressure, cholesterol and diabetes prevent events that cost orders of magnitude more than the drug. A month of a generic statin costs a few dollars. A cardiac admission does not.

Employee pays nothing. Employer pays less. Both sides win on day one.

Yes. Guaranteed issue shelf-rated plans use one published rate per coverage tier, with no age rating, no census collection and no medical questions. Eligibility conditions apply and not every employer qualifies, but for a group that prices badly in the age-rated market, this category is frequently the lowest-cost option available.

The trade is design control. Plan designs are fixed, you can't modify them, and you don't get claims data. For a group whose alternative is a punitive age-rated renewal, that's usually an acceptable trade.

A CHOICE Arrangement offer is affordable when the employee's cost for the lowest-cost silver plan in their rating area, minus the employer's monthly contribution, falls at or below a percentage of household income set annually by the IRS.

That means affordability is a location calculation before it's a contribution calculation. The same dollar contribution can be affordable in one county and unaffordable in the next, because individual market rates are set by rating area. Employers with workers in several rating areas have to run this by location, not once for the company.

The maximum funding rate is the most you can be billed in the plan year. It's contractual. The expected cost is what most groups actually end up paying once surplus is returned, and it's a projection.

When comparing to a fully insured premium, use the maximum. Comparing a guaranteed premium to someone's projection isn't a comparison, it's a sales technique. Look at both numbers, and make sure you know which one you're being shown.

No, and this is where most proposals overstate.

The levers share denominators. Steerage savings and payment integrity savings both apply to some of the same claim dollars. Pharmacy contract savings and specialty drug management overlap. Level funded margin recapture shrinks the base that every other percentage is calculated against.

Add the published ranges naively and you'll get a number in the low thirties as a percentage of plan cost. Combined results in practice land meaningfully below that. Any model that sums the levers is wrong by construction.

No. The structural levers on this page operate on unit price, contract terms and where care is delivered. None of them require reducing headcount, cutting benefits or changing what an employee is entitled to.

Several of them make the employee experience better. A $0 copay formulary lowers what the employee pays at the counter to nothing. A funded out-of-pocket HRA can leave an employee with lower cost sharing on a leaner plan than they had on a richer one. The savings come from the supply side.