A second contract behind the insurance policy
Reinsurance is a contract through which an insurer, the cedent, transfers part of its insurance risk to another insurer. The arrangement can diversify concentrated exposures and expand the amount of business a cedent can support. It ordinarily does not release the original insurer from its obligation to its policyholder. The policyholder’s claim and the cedent’s recovery are separate relationships. [1][2]
That separation explains the central trade-off. An insurer can reduce the size of losses it retains while adding a dependence on contract interpretation, recoverability and the timing of payment from another institution. The coverage limit on a presentation slide is therefore not the same as unconditional cash available for every disaster.
Sharing a percentage versus covering a layer
Under a quota-share contract, the reinsurer takes an agreed percentage of covered premiums and losses; commissions can compensate the cedent for acquisition and servicing costs. Excess-of-loss reinsurance instead responds above a specified attachment point, up to a contractual limit. Coverage may apply per risk, per occurrence or in the aggregate. RenaissanceRe’s 2025 Form 10-K describes both proportional and excess-of-loss business and defines these structures in its glossary. [2]
In a simplified 40% quota share, $100 million of eligible premium and $70 million of covered losses allocate $40 million of premium and $28 million of losses to the reinsurer, before commissions and other contract terms. The cedent keeps $60 million and $42 million respectively. The agreement changes the dollars retained but leaves the loss ratio on this deliberately uniform example at 70% for both parties. It does not magically turn a poorly priced book into an economically sound one.
By contrast, a $40 million excess-of-loss layer attaching at $20 million pays nothing on a $15 million covered loss, $15 million on a $35 million loss and its full $40 million on a $70 million loss. Above exhaustion at $60 million, additional loss again falls outside that layer. The size of the original claim alone cannot identify the recovery without the attachment and limit.
A hypothetical catastrophe tower
Assume one covered occurrence, no exclusions, no coinsurance, full collection and two layers: $40 million in excess of $20 million, followed by $60 million in excess of $60 million. The insurer retains the first $20 million. The first layer covers losses from $20 million through $60 million, and the second from $60 million through $120 million. The recovery is the sum of each layer’s covered slice.
For a $90 million loss, the first layer pays $40 million and the second pays $30 million. The cedent retains $20 million. For a $150 million loss, both layers exhaust, paying $100 million in total, and the cedent retains $50 million: the original $20 million plus $30 million above the tower. A stated $100 million of reinsurance capacity is therefore consistent with substantial retained losses in a large event.
If the first layer is only 80% placed, its $40 million covered slice produces $32 million of recovery. On the $90 million event, the cedent retains $28 million rather than $20 million, even if the upper layer performs in full. This illustrates why an attachment schedule and participation percentages are different pieces of information. The calculations are constructed examples, not a reconstruction of any insurer’s actual program.
The second storm can face different protection
Reinstatement restores used coverage subject to the contract’s conditions and can require an additional premium. An occurrence limit does not establish how many occurrences are covered during the contract period. Aggregate limits, hours clauses defining one event, exclusions and premium adjustments can change the effective protection. Reinsurance purchased by a reinsurer is retrocession; it adds another contractual layer rather than eliminating risk from the system. [2]
Suppose the hypothetical first catastrophe exhausts a $40 million layer and the contract permits one full reinstatement for an additional $4 million. Restoring the layer allows a later covered event to claim against a renewed limit, but the cedent has spent another $4 million for that protection. If no reinstatement is available, the apparent tower is shorter for subsequent losses. A net-loss estimate that subtracts recoveries but ignores reinstatement premiums can understate the economic cost of the event.
RenaissanceRe’s February 3, 2026 release illustrates the distinction in actual reporting. For its category of 2025 large-loss events it separately identified approximately $373.3 million of assumed reinstatement premiums earned and $34.8 million of ceded reinstatement premiums earned. One is revenue on protection it supplied; the other is cost on protection it purchased. These figures are historical accounting components, not the price of a generic catastrophe contract. [3]
Indemnity and catastrophe-bond triggers are different dimensions
Indemnity protection responds to covered losses actually sustained under the contract. A catastrophe bond is a capital-markets funding structure that can use an indemnity trigger, an industry-loss trigger or a parametric trigger tied to physical measurements such as wind speed. It is inaccurate to describe all catastrophe bonds as parametric. The Federal Reserve Bank of Chicago’s primer explains these alternatives. [4]
Imagine an insurer suffers $50 million of claims in a storm but a parametric contract’s specified wind measurement is not reached at the designated location. The payout could be zero despite substantial losses. Conversely, a trigger could pay when that insurer’s own claims are smaller. This mismatch is basis risk. An indemnity trigger reduces that particular mismatch but requires determination of covered loss; a physical trigger may be quicker to assess while matching the financial damage less precisely.
The trade-off is not a universal ranking of products. A public emergency fund seeking rapid post-disaster has a different objective from an insurer seeking reimbursement of a precisely defined portfolio. The relevant question is whether the trigger matches the purpose, not whether one label sounds more sophisticated.
Transferred insurance risk can become credit and liquidity risk
If $70 million is contractually recoverable but only $50 million arrives before policyholder payments are due, the cedent faces a $20 million timing gap. If the remaining claim is ultimately paid, that is pressure; if it is disputed or uncollectible, it may also be an economic loss. Collateral can mitigate some counterparty exposure, but it does not settle every coverage question. RenaissanceRe’s filing identifies exclusions and reinsurer insolvency as limits on protection. [2]
Diversification helps when losses are imperfectly correlated. A severe event affecting several cedents and their common reinsurers can weaken that benefit precisely when claims rise. Conversely, global capital can absorb a concentrated local event more effectively than a small local balance sheet. Both observations can be true without implying that every catastrophe threatens the whole reinsurance market.
Capacity, price and the remaining uncertainty
Reinsurance capacity can support additional primary insurance, but a greater supply of risk capital does not guarantee lower retail premiums. The cedent still faces its own claim costs, expenses, capital requirements and market conditions. In a hypothetical renewal, lower reinsurance pricing accompanied by a higher attachment could mean the insurer pays less while keeping more frequent losses.
Useful evidence includes the actual attachment, limits and participation at renewal; the treatment of multiple events; recovery collection; and how the insured exposure itself has changed. This article concerns contractual risk transfer. The combined ratio and reserve-development article addresses how the resulting insurance activity is reported in earnings. Neither a favorable ratio nor a large reinsurance limit substitutes for the other.
Sources
- NAIC; Reinsurance; current explanatory page checked October 4, 2026SourceBack to text: ↑
- RenaissanceRe; 2025 Form 10-K; year ended December 31, 2025, business, risk factors and glossaryFiling / reportBack to text: ↑1↑2↑3↑4
- RenaissanceRe; 2025 results, large-loss event reconciliation; February 3, 2026SourceBack to text: ↑
- Federal Reserve Bank of Chicago; Catastrophe Bonds: A Primer and Retrospective; Chicago Fed Letter 405, 2018SourceBack to text: ↑