4.5 Program Layering: Retentions, Excess, Umbrella, and Reinsurance
Key Takeaways
- Under a deductible the insurer pays from dollar one, controls the defense, and seeks reimbursement; under a self-insured retention the insured pays first, usually controls defense within the retention, and the policy limits sit in addition to the SIR.
- An umbrella broadens coverage as well as increasing limits and can drop down when underlying limits are exhausted by covered losses, but most forms do not drop down when the underlying carrier is insolvent.
- Excess layers may be following form or standalone; a mid-tower layer with different definitions or notice conditions can create a gap that is covered below it and above it but not within it.
- Quota share reinsurance is proportional and cedes a fixed percentage of premium and losses; excess of loss is non-proportional and attaches above a stated retention to cut off severity.
- A fronting carrier issues admitted paper and cedes the risk back to the captive, charging a fronting fee and requiring collateral such as letters of credit, trusts, or funds withheld.
Quick Answer: A healthcare insurance program is a vertical tower. Under a deductible, the insurer pays the claim from dollar one, controls the defense, and then bills the insured back. Under a self-insured retention (SIR), the insured pays first and generally controls the defense until the retention is exhausted, and the policy limits sit in addition to the SIR. Above the primary layer sit buffer and excess layers, with an umbrella that broadens coverage as well as increasing limits.
Why the Exam Tests Program Structure
Domain 2 task A puts the risk manager in charge of managing a broad, comprehensive insurance program. Items test whether you can read a tower, say which layer pays what, explain why a given retention was chosen, and recognize the structural traps that leave an organization uninsured in the middle of its own program.
Note the authority boundary before the mechanics. The risk manager models, recommends, and explains retention and limit options. The decision belongs to the chief financial officer, the finance committee, and the board, informed by actuarial analysis. An answer choice in which the risk manager unilaterally raises the retention or drops a layer to save premium exceeds the role.
Retention: Deductible vs. Self-Insured Retention
Both are retained risk, and both sit below the insurance. The tested difference is who pays first and who controls the defense.
Deductible. The insurer's obligation attaches at dollar one. The carrier adjusts the claim, appoints and directs defense counsel, pays the loss, and then seeks reimbursement of the deductible amount from the insured. The insured has little control over strategy, and on many forms the deductible is eroding, meaning amounts within it reduce the policy limit rather than sitting beneath it.
Self-Insured Retention (SIR). The insured's obligation comes first. The carrier's duties — including, in most forms, the duty to defend — do not attach until the SIR is exhausted by payments made by the named insured. The organization funds defense and indemnity within the retention, usually administering claims through a third-party administrator (TPA), and it generally directs defense strategy in that band. Policy limits sit in addition to the SIR.
Two consequences follow. First, an SIR demands real infrastructure — claims staff or a TPA, a funding mechanism, and reserve discipline. Second, most forms require the SIR to be satisfied by the named insured itself; payments made by another party can leave the retention technically unexhausted and the excess layer unattached. Self-insurance trusts and captives that fund retentions are covered separately; what matters here is the payment sequence and defense control.
Building the Tower
| Element | What it is | Practical effect |
|---|---|---|
| Retention | Deductible or SIR at the base | Determines who pays first and who directs defense |
| Primary layer | The first insured layer, attaching immediately above the retention | Usually the layer with the duty to defend and the most detailed wording |
| Buffer layer | A modest layer inserted between primary and the first true excess layer | Bought when excess markets will not attach as low as the primary's top |
| Excess layers | Successive layers of limit, each attaching where the one below ends | May be following form or standalone wording |
| Umbrella | Excess over scheduled underlyings that also drops down | Broadens coverage as well as increasing limit |
| Attachment point | The dollar figure at which a layer begins to pay | A layer pays nothing until the loss reaches it |
Following form versus standalone wording is a live risk, not a formality. A following-form excess policy adopts the terms, conditions, and exclusions of the underlying policy, so the tower behaves as one contract. A standalone or manuscript excess policy has its own definitions, exclusions, and notice conditions. When a mid-tower layer defines "claim" differently or imposes a stricter notice condition than the primary, a loss can be covered below it and above it but not in it. Risk managers ask the broker for a wording comparison across the tower and escalate mismatches.
Umbrella versus excess. An excess policy only adds limit. An umbrella adds limit and breadth: it responds excess of the scheduled underlying policies, and it can drop down to act as primary — subject to its own self-insured retention — for a loss that the underlying policies do not cover at all.
Drop-down has a much-tested boundary. Umbrellas drop down when underlying limits are exhausted by payment of covered losses. They generally do not drop down when the underlying limit is uncollectible because the underlying carrier is insolvent; most modern forms state that the insured is deemed to be self-insured for that layer. Insurer financial strength is therefore a live risk management issue, not just a procurement preference.
Per-occurrence versus aggregate limits. The per-occurrence limit caps what a layer pays for one loss. The annual aggregate caps what it pays for the whole policy year. Aggregates erode. A system that absorbs three large professional liability losses in the first eight months can find its primary aggregate exhausted with four months left, at which point the retention effectively sits directly under the next layer or the band is uninsured. Monitoring aggregate erosion against the loss run, and pricing a mid-term reinstatement or a buy-back layer before the aggregate closes, is a core Domain 2 activity.
A Worked Claim Through the Tower
A regional health system carries this professional liability program:
| Layer | Limit | Attaches at | Who pays and controls |
|---|---|---|---|
| Self-insured retention | $2,000,000 per occurrence | $0 | System pays; TPA adjusts; system directs defense within the retention |
| Primary | $3,000,000 occurrence / $9,000,000 aggregate | $2,000,000 | Primary carrier; duty to defend attaches once the SIR is exhausted |
| Buffer | $5,000,000 per occurrence | $5,000,000 | Buffer carrier, following form |
| First excess | $15,000,000 per occurrence | $10,000,000 | Excess carrier, following form |
| Second excess | $25,000,000 per occurrence | $25,000,000 | Excess carrier with consent-to-settle rights |
Total program: $48,000,000 of insurance above a $2,000,000 retention, for $50,000,000 of protection per occurrence.
A birth-injury case settles for $32,000,000. Each layer pays only within its own band:
| Layer | Band | Pays on this claim | Status afterward |
|---|---|---|---|
| Self-insured retention | $0 – $2M | $2,000,000 | Satisfied for this occurrence; resets for the next one |
| Primary | $2M – $5M | $3,000,000 | Occurrence limit exhausted; $6,000,000 of the $9,000,000 aggregate remains |
| Buffer | $5M – $10M | $5,000,000 | Exhausted for this occurrence |
| First excess | $10M – $25M | $15,000,000 | Exhausted for this occurrence |
| Second excess | $25M – $50M | $7,000,000 | $18,000,000 of limit remains |
| Total | $32,000,000 |
Read the aggregate line carefully, because that is where items hide. Two more claims of similar size will close out the primary's $9,000,000 aggregate. Once it is gone, nothing sits between the system's retention and the buffer's $5,000,000 attachment point, so the system self-funds the $2,000,000 to $5,000,000 band on every subsequent loss for the rest of the policy year. That is the finding the risk manager escalates to the finance committee — not after the year closes, but the week the second large reserve posts.
Notice also that the loss pierced four layers. Every one of those carriers has its own notice condition, and the excess carriers had to be put on notice long before the settlement number arrived. Notice up the tower is treated in its own section and is the most common way a structurally sound program still fails.
A health system carries a $1,000,000 per-claim self-insured retention beneath its hospital professional liability tower. A malpractice suit is served with a $750,000 demand. Who funds the defense and who directs it?
A hospital's primary layer exhausts its $10,000,000 annual aggregate in September through payment of covered losses. A new covered occurrence happens in October. Under a typical umbrella written over that primary, what happens?
A health system's captive cedes a fixed 40 percent of every premium dollar and 40 percent of every loss dollar to a reinsurer, across the whole book. This arrangement is best described as:
Reinsurance, Fronting, and Collateral
Reinsurance is insurance for the insurer. The hospital is not a party to it and cannot claim directly against its captive's reinsurer. Two structures dominate:
- Quota share is proportional. The ceding company transfers a fixed percentage of every premium dollar and every loss dollar — a 40% quota share cedes 40% of both. It is used to relieve surplus strain and to buy stability while a program builds credible data.
- Excess of loss is non-proportional. The reinsurer pays only above a stated retention, per risk or per occurrence, up to a limit. It is used to cut off severity, and a healthcare captive typically buys it so a single catastrophic obstetric or neurosurgical verdict cannot impair surplus.
- Aggregate stop-loss caps the ceding entity's total annual retained losses, protecting against an accumulation of ordinary claims rather than one shock loss.
A captive buys reinsurance to dampen volatility, protect capital and surplus, gain capacity beyond what its own balance sheet supports, and satisfy the solvency requirements of its domicile regulator.
Fronting solves a different problem. Contracts, statutes, and lenders frequently demand a policy issued by an admitted, licensed, and rated carrier — a captive often is not. In a fronting arrangement an admitted carrier issues the policy in its own name and cedes the risk back to the captive under a reinsurance agreement. The captive bears the economics; the front lends its paper and files the certificates.
The front is not a bystander. It remains legally liable to the insured and to third parties if the captive cannot pay, so it charges a fronting fee expressed as a percentage of premium and demands collateral for the ceded obligations:
- Letters of credit (LOC) from a bank acceptable to the front, which consume the health system's borrowing capacity.
- Trust agreements, often Regulation 114 trusts, holding assets for the front's benefit.
- Funds withheld or cash deposits.
Collateral is a real cost of alternative risk financing that boards routinely underestimate. It is also usually indexed to loss reserves, so it grows as the program's reserves develop.
How a Retention Level Is Actually Chosen
Retention selection is a financial decision supported by analysis, not a preference:
- Risk appetite and risk tolerance. Appetite is how much risk the organization is willing to accept in pursuit of its mission; tolerance is the acceptable variation around that. Both are board-set.
- Loss history and predictability. High-frequency, low-severity losses are predictable and cheap to retain. Low-frequency, high-severity losses are volatile and should be transferred.
- Credibility of the data. A small system with thin loss experience cannot support the same retention as a system with two decades of credible data.
- Balance sheet and cash flow. Days cash on hand, debt covenants, and the rating agencies' view of self-insured liabilities all constrain how much loss the organization can absorb in a bad year.
- Actuarial confidence level at which the retained layer will be funded. The choice of confidence level and the discounting of reserves are treated in the actuarial section; what matters structurally is that a higher retention requires a larger funded reserve.
- Market pricing. The broker prices the premium credit for each increment of retention bought. If buying the retention down from $2,000,000 to $1,000,000 costs more than the expected losses in that band plus a risk margin, the organization keeps the higher retention.
- Collateral and administrative cost, including TPA fees and letters of credit.
Exam Traps on Program Structure
- Reversing deductible and SIR. Insurer pays first and bills back equals deductible. Insured pays first and generally controls defense equals SIR.
- Assuming limits always sit above the retention. They sit above an SIR, but an eroding deductible reduces the limit.
- Letting someone else fund the SIR. Most forms require satisfaction by the named insured; a payment from a parent or another insurer can leave the excess unattached.
- Treating umbrella and excess as synonyms. Excess adds limit only; umbrella adds limit and breadth and can drop down.
- Expecting drop-down on carrier insolvency. Drop-down follows exhaustion by covered loss, not an underlying carrier's failure.
- Ignoring aggregate erosion until renewal. The uninsured band opens mid-year, silently.
- Assuming a following-form tower. Standalone excess wording can open a mid-tower gap in notice or definitions.
- Thinking reinsurance protects the hospital. It protects the insurer or captive; the hospital has no privity with the reinsurer.