Most employers treat the move to a self-funded health plan as a decision they will make when the time is right. The problem is that the time is right about eighteen months before anyone thinks to ask. By the point a growing company crosses fifty employees and starts shopping alternatives to its fully insured renewal, the single asset that determines whether self-funding is even on the table has to already exist. That asset is claims history, and it cannot be bought, borrowed, or reconstructed after the fact.

This is the sequencing trap that catches employers between forty and sixty lives. The compliance clock and the underwriting clock run on different schedules, and almost nobody lines them up.

Key Takeaways
  • The ACA Applicable Large Employer threshold of 50 full-time and full-time-equivalent employees is a predictable event you can forecast from your headcount trend, not a surprise.
  • Self-funded and captive underwriters price your plan off 12 to 24 months of your own claims experience, which fully insured small groups usually cannot get from their carrier.
  • Level-funded plans are the practical bridge because they return monthly claims and experience reports that build the data record self-funding will later require.
  • KFF data shows self-funding jumps from 27 percent of covered workers at firms with 10 to 199 employees to 80 percent at firms with 200 or more. The gap is largely a data-readiness gap.
  • The move worth planning is not the switch to self-funding. It is the earlier switch to a funding model that starts generating claims data, ideally at 35 to 40 lives.

The threshold you can see coming

The Affordable Care Act defines an Applicable Large Employer, or ALE, as one that averaged at least 50 full-time and full-time-equivalent employees during the prior calendar year. Full-time means an average of 30 or more hours per week. Full-time equivalents come from your part-time population: add up all the hours worked by part-timers in a month, cap each person at 120, divide the total by 120, and that count gets added to your full-time headcount. A company with 44 full-timers and enough part-time hours to equal 8 FTEs is an ALE at 52, even though the payroll system shows 44 salaried people.

Crossing that line turns on the employer shared responsibility provisions under Section 4980H. For 2026 the IRS has set the two penalty amounts at levels worth knowing before you get near them. The 4980H(a) penalty, which applies when an ALE fails to offer minimum essential coverage to at least 95 percent of full-time employees and at least one of them buys subsidized Marketplace coverage, runs $3,340 per full-time employee for the year, minus the first 30. That works out to $278.33 per month. The 4980H(b) penalty, which applies when coverage is offered but is either unaffordable or fails minimum value, runs $5,010 per affected employee for the year, or $417.50 per month. Affordability for 2026 is pegged at 9.96 percent of household income for the lowest-cost self-only option.

The point of reciting those numbers is not the penalties themselves. It is that you can predict the exact year you become an ALE. Headcount growth is one of the few things about a benefits program you can graph in advance. If you are hiring on a steady curve, you know roughly which quarter takes you past 50. That predictability is the whole opportunity, because the other clock that matters is one you cannot rush.

The data wall most fully insured employers hit

Here is what surprises employers when they first ask their broker about self-funding: they cannot get their own claims data. A fully insured group under roughly 100 lives is typically community rated or pool rated. The carrier prices the group off the block of business it belongs to, not off its individual experience, and it has no obligation to hand back a detailed claims report. You get a renewal percentage and a rate sheet. You do not get the paid-claims-by-month, high-cost-claimant, and diagnosis-category detail that any self-funded quote depends on.

So when a 52-life employer decides it wants to explore self-funding, the first request from a stop-loss underwriter is 12 to 24 months of claims experience. And the honest answer is often that the data does not exist in usable form, because the prior arrangement never produced it. You cannot underwrite what you cannot see, and you cannot retroactively create a claims record for a period that has already passed under a plan that never tracked it.

This is the wall. It is not a pricing wall or a compliance wall. It is an information wall, and it sits directly between a growing employer and the funding models that would save it money at scale.

What self-funded and captive underwriters actually ask for

When you move from fully insured to any risk-bearing structure, the underwriting question changes from "what block do you belong to" to "what does your group actually cost." Stop-loss carriers, which sell the catastrophic coverage that makes self-funding safe, want to see the shape of your risk. A typical submission for a self-funded or captive quote asks for:

Roughly 15 to 20 percent of an enrolled population usually drives about 80 percent of claims cost, so that high-cost claimant report is the document underwriters read first. Without a real claims history, a stop-loss carrier either declines to quote or loads the rate heavily to cover the uncertainty, which erases the savings that made self-funding attractive in the first place. The data is not paperwork. It is the price.

Model your funding options before your headcount forces the question

The Health Funding Projector lets you compare fully insured, level-funded, and self-funded scenarios side by side using your real numbers, so you can see when a data-building move starts paying off.

Level-funded plans are the data bridge, not just the cheaper option

Most coverage of level-funded plans frames them as a middle option: a little more risk than fully insured, a little less than self-funded, with a fixed monthly payment that smooths out cash flow. All true. The part that gets missed is that a level-funded plan produces the thing a fully insured plan withholds. Because a level-funded arrangement is a self-funded plan wrapped in stop-loss and a level monthly bill, the administrator tracks and reports your claims. You get monthly experience reports, high-cost claimant flags, and a surplus-or-deficit position at year end.

In other words, the move to level-funded is the moment your group starts generating an underwritable claims record. That reframes the whole decision. A level-funded plan is not only a way to possibly save money now. It is the vehicle that makes the eventual move to full self-funding or a captive quotable at all, because it lays down 12 to 24 months of the exact experience data those structures require.

The KFF 2025 Employer Health Benefits Survey shows how the market already behaves this way, even if employers do not describe it in these terms. Among firms with 10 to 199 workers, 27 percent of covered workers are in self-funded plans, while 37 percent are in level-funded plans. At firms with 200 or more workers, self-funding rises to 80 percent. Read that as a pipeline rather than three separate choices. Level-funded is where mid-market employers accumulate the experience and the comfort that let them graduate to self-funding as they grow. The 37 percent in level-funded plans are, in many cases, the self-funded employers of two years from now, whether they know it yet or not.

The cost of arriving without data

It helps to put rough numbers on what the data wall costs. Consider a hypothetical 55-life employer paying about $9,300 per employee per year on a fully insured plan, or roughly $511,000 in annual premium. Suppose a self-funded structure priced off a clean two-year claims record would land near $445,000 in expected total cost, a savings of about 13 percent. That is the prize.

Now remove the data. With no credible claims history, a stop-loss underwriter has to price the unknown. The common response is a defensive load on the stop-loss layer and a conservative expected-claims assumption, which can add 10 to 15 percent to the projected cost. The self-funded quote comes back at or above the fully insured number, so the employer reasonably concludes self-funding does not pay and stays put. The savings were real. They were just locked behind a data record that did not exist.

The same employer, having spent the prior two years on a level-funded plan, walks in with monthly experience reports showing where the claims actually went. The underwriter prices off real numbers instead of a worst-case assumption, the quote reflects the group's true risk, and the 13 percent is suddenly reachable. Nothing about the underlying population changed between those two scenarios. The only variable was whether the claims history existed. That is the entire argument for starting early: the data is worth real money, and it accrues only with time.

A timing playbook that works backward from 50

If the ALE date is predictable and the data takes 12 to 24 months to build, the planning move is simple arithmetic run in reverse. Start from the quarter you expect to cross 50 full-time and full-time-equivalent employees. Subtract the underwriting look-back. Land on the date you should already be generating claims data. For most growing employers that puts the right move to a data-producing funding model somewhere around 35 to 40 lives.

Working it as a sequence:

  1. At 30 to 35 lives, forecast your ALE date. Graph your headcount over the last eight quarters and project it forward. Include the FTE math, not just salaried heads, because part-time hours can move the date earlier than you expect.
  2. At 35 to 40 lives, move to a level-funded plan. The immediate reason is cost and cash-flow control. The strategic reason is that this is when the claims meter starts running. Every month under level funding is a month of experience data you will be able to hand an underwriter later.
  3. Capture and keep every report. Save the monthly experience reports, the high-cost claimant lists, and the year-end settlement. This is your future self-funded submission, assembled in real time.
  4. At 48 to 52 lives, shop the risk-bearing structures. Now you have the trailing data a stop-loss carrier or captive wants, and you can get a real quote instead of a declined one. The ALE compliance obligations arrive the same year, so you handle funding strategy and mandate compliance in one coordinated cycle rather than two scrambles.

The employers who struggle are the ones who skip step two. They stay fully insured until the renewal pain at 52 finally pushes them to look, then discover the data wall and lose another 18 months building the record they could have been building all along.

What this means for your renewal cycle

Renewals are where this either pays off or hurts. A fully insured employer approaching the threshold is negotiating blind, because the carrier holds the only detailed view of the group's cost. A level-funded employer walks into the same renewal holding its own claims report, which changes the conversation. You can see whether a bad year was one large claim or a broad trend. You can decide whether to adjust plan design based on where the dollars actually went. And when the time comes to solicit self-funded or captive quotes, you are handing over a clean two-year record instead of apologizing for not having one.

The compliance side lines up too. The year you become an ALE is the year the 4980H affordability and offer requirements attach. Handling that transition while you already have visibility into your own claims is far easier than doing it during a panic switch. Reviewing your funding model and your ALE status in the same planning cycle, ideally the year before you cross the line, turns two separate fire drills into one deliberate move.

Related reading

Frequently Asked Questions

How much claims history do I need to get a self-funded or captive quote?

Most stop-loss underwriters ask for 12 to 24 months of paid claims experience, separated into medical and pharmacy, along with a high-cost claimant report. Twenty-four months gives the cleanest read because it lets an underwriter see whether a high-claim period was a one-time event or a trend. Twelve months is often workable but tends to draw a more cautious rate.

Why can't I just get claims data from my current fully insured carrier?

Fully insured groups under roughly 100 lives are usually community or pool rated, which means the carrier prices you off a larger block of business rather than your own experience. Because your individual claims do not drive your rate, the carrier has no obligation to produce a detailed report, and most do not. This is the core reason a data-producing arrangement has to come first.

Is a level-funded plan risky for an employer around 40 employees?

Level-funded plans are built for exactly this size. The employer pays a fixed monthly amount, and stop-loss coverage caps exposure on both individual large claims and total annual claims. If claims come in low, the employer may receive a year-end surplus. If they run high, the stop-loss layer absorbs the excess. The main gain for a growing employer is that the arrangement produces the claims data a future self-funded move will require.

What happens if I wait until after I cross 50 to start?

You can still move, but you will likely spend another 12 to 24 months building the claims record before self-funded or captive underwriting becomes favorable. Waiting does not remove the data requirement. It just starts the clock later, which delays the point at which the savings from self-funding become real.

Does the FTE calculation really change when I become an ALE?

Yes, and it catches employers off guard. Full-time-equivalent employees are calculated from part-time hours: total the monthly hours of part-timers, cap each individual at 120, and divide by 120. Those FTEs are added to your full-time count. A company that looks like it is safely under 50 on salaried headcount can cross the ALE threshold once part-time hours are included, so run the full calculation when you forecast your date.

This article is educational and does not constitute legal or tax advice. Consult your own benefits, legal, and tax advisors for guidance specific to your situation.

Sources referenced: IRS employer shared responsibility provisions under Section 4980H, 2026 indexed penalty amounts and affordability percentage, irs.gov. KFF 2025 Employer Health Benefits Survey, self-funding and level-funding prevalence by firm size, kff.org.