Key Findings
- 0141% of self-insured employers carry specific stop-loss attachment points set above the actuarially optimal level, paying excess premium without proportionate protection
- 02Average stop-loss overpayment for mid-market employers (500–5,000 lives): $180 per employee per year
- 03Data-driven attachment point modeling using actual claims history recovers $90K–$450K annually depending on plan size
- 04The right stop-loss decision requires three years of claims run-out data — not national benchmark tables — to set attachment points accurately
Stop-loss insurance exists to protect self-insured employers from catastrophic individual claims. A single employee with a cancer diagnosis, a premature birth, or a major trauma can generate $500,000 or more in claims in a single year. Stop-loss coverage — specifically, specific stop-loss with an individual attachment point — caps the employer's liability for any single member, transferring excess risk to a carrier.
The concept is sound. The execution, for 41% of self-insured employers, is not.
The problem is attachment point selection. Most employers set their specific deductible based on gut feel, broker recommendation, or national benchmark tables — not on their own claims history. When the attachment point is set too high, the employer pays premiums for catastrophic protection that statistical analysis of their own data shows they rarely need at that threshold. That's overpayment.
The Landscape
Self-insurance has grown significantly over the past decade. Today, 65% of workers with employer coverage are on self-funded plans. As more mid-market employers have moved to self-insurance, the stop-loss market has grown correspondingly — and the pricing opacity has followed.
Stop-loss carriers compete aggressively on premium at the front end of a relationship, then adjust at renewal based on claims experience. An employer who had a bad year sees sharp premium increases. An employer who had a good year may not see commensurate decreases. The asymmetry is predictable and largely unavoidable — but the baseline attachment point decision is within the employer's control.
Two additional market forces have shifted in the last two years. First, specialty drug costs have increased the frequency of high-dollar individual claims — GLP-1s, oncology biologics, and gene therapies are producing more $200,000+ annual claims per 1,000 covered lives than actuaries expected five years ago. Second, stop-loss carriers have responded by tightening exclusions and increasing rates, making the decision about which risks to transfer and at what level more consequential than it was in a lower-drug-cost environment.
What the Data Tells Us
Three data points drive the stop-loss attachment point decision:
Your historical large claim frequency. How many claims has your plan had above $100,000, $200,000, and $500,000 in each of the last three years? This tells you the actual probability of a catastrophic claim, specific to your population. National benchmark tables approximate this — but your actual experience is more predictive.
Your plan's risk tolerance. A cash-rich employer with stable margins can absorb more variance and set a higher attachment point (lower premium). An employer with thin margins or volatile revenue needs more protection and should set a lower attachment point (higher premium). The right answer is not universal.
The marginal cost of attachment point reduction. Stop-loss carriers price each $25,000 reduction in specific deductible as an incremental premium. Often, the marginal cost of buying protection from $300,000 down to $200,000 is much more favorable than buying protection from $200,000 down to $150,000. The pricing curve is not linear, and the value points are not obvious without running the numbers.
Our analysis of mid-market employer stop-loss programs shows that 41% of employers have specific attachment points set above the level that minimizes total plan cost (premium plus expected retained liability). The average overpayment is $180 per employee per year — not dramatic on a per-member basis, but significant in aggregate. For a 500-life plan, that's $90,000 annually. For a 2,500-life plan, it's $450,000.
The Framework
Use this decision framework for your next stop-loss renewal:
Step 1: Pull three years of claims run-out. Ask your TPA for a report of all individual claims that exceeded $50,000 in each of the last three plan years, with 12-month run-out. You need actual paid amounts, not incurred estimates.
Step 2: Calculate your own large-claim probability. For each threshold ($100K, $150K, $200K, $300K, $500K), calculate how many claims you've had above that threshold in each of the last three years, and what the maximum individual paid amount has been. This is your empirical starting point.
Step 3: Model the premium vs. retention trade-off. For the next plan year, ask your stop-loss broker to quote three specific deductible levels: your current deductible, your current deductible plus $50,000, and your current deductible minus $50,000. Compare the premium differential to your historical probability of claims above each threshold. The math will often reveal that you're paying meaningful premium for coverage at a threshold you've never hit.
Step 4: Evaluate laser provisions carefully. A "laser" is a stop-loss carrier's right to exclude a specific known high-cost claimant or increase their individual attachment point. If a carrier is lasering a chronic high-cost member, understand the financial exposure before accepting the contract. The premium discount may not be worth the retained liability.
Step 5: Benchmark your stop-loss carrier's loss ratio. A stop-loss carrier with a low loss ratio relative to the market is likely overcharging for the risk. Your broker should be able to provide competitive market data on loss ratios by carrier and segment.
Risks and Trade-offs
The tail risk is real. Increasing your specific deductible to reduce premium means accepting more variance. In a year with an unexpected catastrophic claim — a preterm twin delivery, a transplant case, an emerging oncology diagnosis — the retained liability can be material. Model the worst-case scenario before accepting a higher deductible.
Specialty drugs create new cliff risks. Gene therapies priced at $2M–$4M per patient are now entering the market. Traditional stop-loss contracts may exclude or limit coverage for these treatments. Review your contract language specifically for specialty and gene therapy carve-outs before renewal.
Aggregate stop-loss is a separate decision. Specific stop-loss protects against individual catastrophic claims. Aggregate stop-loss protects against a year of unusually high utilization across the plan. Both are relevant but priced differently. Don't optimize specific without also reviewing aggregate corridor levels.
Getting Started
Data needed: Three years of claims run-out by individual member, large claim listing above $50K threshold, current stop-loss contract and premium by layer.
Who to involve: TPA (for data), stop-loss broker (for market quotes), actuary or benefits consultant (for modeling), CFO (for risk tolerance decision).
Timeline: Begin the attachment point modeling analysis at least 90 days before your stop-loss renewal date. Carrier negotiations require time, and you need comparison quotes from at least three carriers.
How to measure success: Track your actual stop-loss premium as a percentage of total health spend annually. Compare your attachment point to the highest individual claim paid each year. If the gap between your attachment point and your highest claim is consistently large, you are paying for protection you're not using.
Stop-loss optimization is one of the few cost control initiatives that doesn't require changing benefits, reducing network access, or asking employees to behave differently. It's a financial structuring decision — one that most employers are getting wrong by defaulting to inertia rather than analysis.
"Most employers set their stop-loss attachment point once and renew it forever. Running the actual math on your own claims history usually reveals you're paying for protection you don't need — at the level you're buying it."
See how these findings apply to your plan.
We run this analysis on actual claims data for self-insured employers. Your dedicated partner delivers these insights every month.