Attachment Point

Attachment Point Where Stop-Loss and Health Reinsurance Start to Pay

In insurance, the attachment point marks where a payer's own liability for claims ends and its stop-loss insurer, reinsurer or a public risk pool takes over. Below it the employer, insurer or sickness fund pays every claim itself; above it the other party shares the cost, often only up to a cap and at an agreed rate. The general excess-of-loss meaning is the same across insurance lines, and its use in health, from the payment rule and pricing to US regulatory floors, public reinsurance pools and the incentives the choice creates, is the focus here.

The excess layer of retention, coinsurance and cap

An attachment point is the lower edge of an excess layer. The National Association of Insurance Commissioners (NAIC) model act on stop-loss defines it as the claims amount incurred by an insured group beyond which the insurer incurs a liability for payment, and the NAIC white paper on stop-loss also calls the specific attachment point the "retention level". US regulations for the reinsurance programme of the Affordable Care Act (ACA) add a coinsurance rate, the share reimbursed between the attachment point and the cap, and a reinsurance cap, above which claims are not eligible.

For one covered person in one benefit year, a layer with all three parameters pays the amount below.

$$R = c \times \left[\min(C, U) - A\right]_+$$

where $R$ is the reimbursement to the payer, $C$ is the person's eligible claims for the year, $A$ is the attachment point, $U$ is the cap, $c$ is the coinsurance rate, and $[x]_+$ means $x$ when $x$ is positive and zero otherwise. The payer keeps $C - R$. With $c = 1$ and no cap, retained claims never exceed $A$; coinsurance below 100%, a cap, a laser, exclusions or claims paid outside the contract year can push them higher, and the premium comes on top, so the attachment point is not a cap on the payer's cost.

Specific and aggregate attachment points

US stop-loss policies bought by self-funded employers usually carry two attachment points. The NAIC white paper describes a specific one that protects against claim severity and an aggregate one that protects against claim frequency.

• Specific (individual) attachment point. A threshold on one covered person's claims in the contract year; the federal Employee Benefits Security Administration (EBSA) gives USD 500,000 per enrollee as an example.
• Aggregate attachment point. A threshold on the employer's total retained claims for the year, usually a percentage of expected claims (EBSA's example is 125%), net of specific recoveries; see aggregate stop-loss.
• Lasered attachment point. The NAIC white paper describes a "laser specific", a higher attachment point for individuals with known high-cost conditions or other identified risk factors, offered as a condition of cover or for a lower premium.

Choosing the level by weighing premium against retained risk

The NAIC white paper states that the attachment points determine how much risk the employer retains and transfers, and that what it will pay for lower ones depends on the risk it can afford and what the state regulator allows. Expected reimbursement falls steeply as the attachment point rises, but variability falls more slowly, because the rare very large claims sit at the top of the distribution. Fixed money attachment points also drift with claims inflation. As costs rise, every claim already above a fixed threshold grows by its full increase while the threshold stays put, and more people cross it, so the expected cost of the layer grows faster than average claims.

Worked example with three specific attachment points for one employer

The figures are illustrative. A self-funded employer covers 1,000 members. Each member's annual eligible claims are below USD 50,000 with probability 0.990, USD 75,000 with probability 0.006, USD 150,000 with probability 0.003 and USD 500,000 with probability 0.001, and members' claims are treated as independent. The employer compares specific attachment points of USD 50,000, USD 100,000 and USD 200,000, each with full reimbursement and no cap, priced at expected reimbursement plus a 30% loading.

1. Expected reimbursement per member. For a discrete distribution of claims, it is calculated as follows.

$$E[R] = \sum_{k} p_k \times (C_k - A)_+$$

where $E[R]$ is the expected reimbursement for one member, $p_k$ is the probability of claim level $C_k$, and $A$ is the attachment point. The expected reimbursement at an attachment point of USD 50,000 is as follows.

$$E[R] = 0.006 \times 25{,}000 + 0.003 \times 100{,}000 + 0.001 \times 450{,}000 = 150 + 300 + 450 = 900$$

where each term is a probability times the amount above USD 50,000. At USD 100,000 the expected reimbursement is 0.003 × 50,000 + 0.001 × 400,000 = 550 per member, and at USD 200,000 it is 0.001 × 300,000 = 300 per member.

2. Variability of the transferred claims. Per member, the variance is as follows.

$$Var(R) = 0.006 \times 25{,}000^2 + 0.003 \times 100{,}000^2 + 0.001 \times 450{,}000^2 - 900^2 = 235{,}440{,}000$$

where $Var(R)$ is the variance of one member's reimbursement at the USD 50,000 attachment point. For 1,000 independent members the standard deviation of total reimbursement is the square root of 1,000 times this, about USD 485,222. The same steps give USD 408,898 at USD 100,000 and USD 299,850 at USD 200,000.

3. Results for the group.

4. Claims inflation. If every claim level rises by 10% (the USD 75,000 level becomes USD 82,500, still below the attachment point), expected reimbursement at the USD 100,000 attachment point becomes 0.003 × 65,000 + 0.001 × 450,000 = 645 per member, a rise of about 17.3%.

5. What the numbers show. Raising the attachment point from USD 50,000 to USD 200,000 cuts expected reimbursement by two thirds and the illustrative premium by USD 780,000, but the standard deviation of what is transferred by only about 38%. The layer given up, from USD 50,000 to USD 200,000, has expected reimbursement of USD 600,000 and a standard deviation of USD 236,411, a ratio of 0.39, against 1.00 for the layer above USD 200,000. The lower layer is mostly spending that recurs from year to year, prepaid with a loading, while the top layer is almost pure catastrophe protection; the employer now keeps up to USD 200,000 per high-cost member.

Regulatory floors on stop-loss attachment points in the United States

The NAIC white paper observes that stop-loss with low enough attachment points works like group health insurance without its regulation, and that many states set thresholds so that employers stay truly self-funded. EBSA's Technical Release 2014-01 warns that low attachment point stop-loss can also create adverse selection in the risk pool.

The NAIC Stop Loss Insurance Model Act, adopted in 1995 and amended in 1999, bars a policy with any of the following.

• an annual specific attachment point below USD 20,000;
• an annual aggregate attachment point, for groups of 50 or fewer, below the greatest of USD 4,000 times the number of group members, 120% of expected claims, or USD 20,000;
• an annual aggregate attachment point, for groups of 51 or more, below 110% of expected claims.

The act lets the insurance commissioner amend these amounts with reference to medical prices, and the floor that applies to a given policy is set by the law of the state that adopts it.

Attachment points in public reinsurance and high-cost risk pools

Public reinsurance writes the attachment point into regulation. Under 45 CFR 153.230, each ACA transitional reinsurance payment was the coinsurance rate times an issuer's claims costs for an enrollee in the year between the attachment point and the cap.

The ACA transitional reinsurance programme (2014 to 2016). The programme covered individual-market plans, and the Department of Health and Human Services (HHS) set its parameters in annual payment notices, shown in the table below.

HHS lowered the 2014 attachment point because it had overestimated covered claims costs, so it could pay more for high-cost enrollees without raising the contribution rate. It kept the cap at USD 250,000 to avoid interfering with commercial reinsurance, which typically attaches in that range. From the 2015 notice, if contributions for a year exceeded requests for payment, that year's coinsurance rate was to be raised, up to 100%.

The German risk pool. Section 268 of the Social Code Book V adds a risk pool (Risikopool) to the risk structure equalisation (Risikostrukturausgleich) between statutory sickness funds. When a fund's eligible expenditure for one insured person in a year exceeds the threshold, 80% of the amount above it is equalised through the pool. The statute sets the threshold at EUR 100,000, to be uprated in later years by the annual change in eligible expenditure per insured person, and pooled expenditure is left out of the standardised expenditure used to calculate the funds' allocations.

Risk selection, risk sharing and incentives for cost control

An attachment point trades protection for payers against their responsibility for costs. Health economists describe the payer's incentive with the "power" of a payment system, the share of the marginal unit of spending that the plan bears. Below the attachment point and above any cap, power is 1; inside the layer it falls to one minus the coinsurance rate, so a lower attachment point or higher coinsurance rate weakens the payer's incentive to manage the costs of people likely to cross it, a form of moral hazard on the part of the protected payer.

The offsetting gains come from risk and selection. Simulating ACA Marketplace enrollees, Layton, McGuire and Sinaiko found that a large part of an insurer's random risk comes from a few extraordinarily high-cost individuals, so reinsurance cut insurer risk sharply while weakening incentives for only a very small share of enrollees. Geruso and McGuire found that a simple reinsurance system scored better on fit, power and balance than the Marketplace risk adjustment scheme. Van Barneveld and colleagues showed that risk sharing reduces incentives for selection and for efficiency alike, and found that outlier risk sharing, which pays above a threshold, gave poorer trade-offs than other forms. A lower attachment point therefore reduces the reward for cream skimming, at the cost of weaker cost control.

In a budget impact analysis from a payer's perspective, a very expensive therapy for a few people falls largely in the stop-loss or reinsurance layer, while the same spending spread across many members stays with the payer unless it pushes total claims past an aggregate attachment point. The protection lasts only for the contract year. The NAIC white paper notes that stop-loss is underwritten again at renewal, when the insurer may raise the premium or laser known high-cost individuals, so a multi-year analysis should not assume unchanged terms. From a health system or societal perspective the reimbursement is a transfer between payers; only the administration and capital costs within the loading, and any change in care, are real resource effects.

Attachment point compared with deductible, risk corridor and risk adjustment

Several mechanisms set a threshold on health costs, but each sits between different parties. The table shows who carries the costs below each threshold.

Risk adjustment deals with predictable differences in risk, while reinsurance attachment points deal mainly with unpredictable large claims, which is why systems often use both.

Sources

• Geruso M, McGuire TG. Tradeoffs in the design of health plan payment systems: fit, power and balance. Journal of Health Economics. 2016;47:1-19. doi:10.1016/j.jhealeco.2016.01.007
• Layton TJ, McGuire TG, Sinaiko AD. Risk corridors and reinsurance in health insurance marketplaces: insurance for insurers. American Journal of Health Economics. 2016;2(1):66-95. doi:10.1162/ajhe_a_00034
• National Association of Insurance Commissioners. Stop Loss Insurance Model Act (Model 92). Adopted 1995, amended 1999. https://content.naic.org/sites/default/files/model-law-092.pdf
• National Association of Insurance Commissioners. Stop Loss Insurance, Self-Funding and the ACA. White paper. 2015. https://content.naic.org/sites/default/files/inline-files/SLI_SF.pdf
• Sozialgesetzbuch Fünftes Buch (SGB V), section 268 Risikopool. Text as accessed 2 October 2026. https://www.gesetze-im-internet.de/sgb_5/__268.html
• US Department of Health and Human Services. Patient Protection and Affordable Care Act; HHS Notice of Benefit and Payment Parameters for 2015. Federal Register. 2014;79:13744 (11 March 2014).
• US Department of Health and Human Services. Patient Protection and Affordable Care Act; HHS Notice of Benefit and Payment Parameters for 2016. Federal Register. 2015;80:10750 (27 February 2015).
• US Code of Federal Regulations. Title 45, part 153, sections 153.20 and 153.230 (Standards related to reinsurance, risk corridors, and HHS risk adjustment under the Affordable Care Act). eCFR text as accessed 2 October 2026. https://www.ecfr.gov/current/title-45/subtitle-A/subchapter-B/part-153
• US Department of Labor, Employee Benefits Security Administration. Guidance on State Regulation of Stop-Loss Insurance. Technical Release 2014-01. 6 November 2014. https://www.dol.gov/agencies/ebsa/employers-and-advisers/guidance/technical-releases/14-01
• van Barneveld EM, Lamers LM, van Vliet RCJA, van de Ven WPMM. Risk sharing as a supplement to imperfect capitation: a tradeoff between selection and efficiency. Journal of Health Economics. 2001;20(2):147-168. doi:10.1016/S0167-6296(00)00077-1