A reinsurance treaty wording is the contract between cedent and reinsurer. The Munich Re wording reads differently than the Swiss Re wording reads differently than the Hannover Re wording - same risk, same layer, same effective date, three different clause architectures and three different cession-mechanic implications. In 2026, the treaty broker or cedent's reinsurance accounting director uses AI-assisted markup to identify the cession-mechanics differences, the loss-occurrence definitions (the 72-hour Named Storm Hours Clause vs. the 168-hour Earthquake Hours Clause), the Ultimate Net Loss definition, the Original Conditions clause, the Errors-and-Omissions clause, the Sunset clause, the Reinstatement language, and the AI-driven UW representations that the reinsurer is now adding to the slip in 2026. This lesson takes a draft treaty wording (quota share on commercial property plus cat-XOL on Tier-1 wind), produces an AI-assisted markup of all critical clauses, compares Munich Re vs. Swiss Re vs. Hannover Re wording on the same risk, and produces the cedent's negotiating brief. The headline rule: treaty wording is where reinsurance economics live or die; AI accelerates markup and comparison but the credentialed treaty broker and reinsurance accounting director own every signature, every redline, and every renegotiation that the wording produces over the seven-year sunset tail.

The Treaty Architecture - Quota-Share Plus Cat-XOL

Quota-share treaty: the cedent cedes a fixed percentage of premium and losses on a defined book; the reinsurer accepts the same percentage. Example: 30% quota share on the carrier's commercial-property book. The reinsurer takes 30% of premium plus 30% of losses and pays a ceding commission (typically 28-32% of ceded premium) back to the cedent to cover the cedent's acquisition and administration costs. Quota-share is proportional reinsurance - the reinsurer's exposure mirrors the cedent's exposure proportionally.

Cat-XOL treaty: the cedent retains losses up to the attachment point; the reinsurer pays losses above attachment up to the limit; the reinsurer's cession is non-proportional. Example: $50M xs $100M cat-XOL for Tier-1 wind (the treaty pays $50M between $100M and $150M of gross cat loss). Cat-XOL applies per catastrophic event, with "event" defined by the hours clause.

The combined structure on Heartland Mutual's book. Heartland has $42M in net retained TIV on the commercial-property book after per-risk excess. Per-risk excess covers individual large risks at a $25M per-risk limit (with facultative cession on three buildings exceeding that limit, per the facultative-cession lesson). The quota-share treaty takes 30% on the net retention after per-risk excess - meaning the reinsurer participates pro-rata on the carrier's net commercial-property book. The cat-XOL treaty covers catastrophic events that aggregate multiple risks across a 72-hour wind window or a 168-hour earthquake window. Each layer has its own wording; the combined structure requires careful coordination so that gross-loss-after-quota-share triggers the cat-XOL attachment correctly.

The per-event vs. aggregate distinction. Cat-XOL treaties can be per-event (each catastrophic event triggers a fresh layer with limit and reinstatement) or aggregate (annual aggregate limit across all events). The 2026 Heartland placement is per-event with 2x reinstatement at additional premium - meaning the layer can pay out three times in the policy year for three distinct catastrophic events. Reinsurer-favorable structures shift toward aggregate with fewer reinstatements; cedent-favorable structures shift toward per-event with more reinstatements.

Named Storm Hours Clause - 72-Hour

The Named Storm Hours Clause defines what constitutes a single loss occurrence for hurricane and named-storm events. The standard 72-hour clause: all losses arising directly or indirectly from a single named storm during a 72-hour period constitute one loss occurrence. The cedent selects the 72-hour window beginning at the time of the cedent's choice; the window includes all losses regardless of geographic spread within the named-storm's affected zone. The reinsurer's cession applies to the entire occurrence so defined.

The cedent's window selection. If Hurricane Ian-style landfall produces 96 hours of damage spanning Florida west coast inland to central Florida, the cedent picks the 72-hour window that captures the maximum loss. The selection is strategic - picking the wrong window can leave $50-150M of losses outside the occurrence and uncollectable from the treaty. The cedent's reinsurance accounting director or treaty broker makes the call with the chief actuary's input.

The clause variations 2026. Some reinsurers tighten to 60-hour or 48-hour clauses in hard markets; the cedent prefers 168-hour or unlimited duration. Negotiation: cedent argues longer hours capture more losses in a single occurrence (favorable for cession); reinsurer argues shorter hours limit per-event exposure and avoid retroactive aggregation of pre- and post-event damage. Industry standard 2026: 72-hour for Named Storm plus 168-hour for Earthquake. Tropical storms below hurricane intensity may have a separate hours clause (96-hour is common for tropical-storm-only events).

Compound events. Some 2026 treaty wordings address compound events (Hurricane plus inland flooding plus tornado outbreaks within the named-storm window). Standard treatment: all damage within the 72-hour window from the named storm aggregates into a single occurrence, including secondary perils like flood and tornado generated by the named storm. The reinsurer-favorable variant separates flood as a distinct peril; the cedent-favorable variant aggregates all damage.

Earthquake Hours Clause - 168-Hour

Earthquake events with aftershocks can extend over a week. The industry-standard hours clause is 168-hour (7 days). The cedent selects the window beginning at the cedent's choice; all earthquakes and aftershocks during the 168-hour window aggregate into a single occurrence. The reinsurer's cession applies to the aggregate.

2026 specific concerns. Induced seismicity - earthquakes caused by wastewater injection in Oklahoma, Texas, and Kansas - raises questions about whether they fall under the earthquake hours clause or are separate occurrences. Reinsurer wordings increasingly address explicitly: some exclude induced seismicity entirely; others include it but with a separate sub-limit; some treat it as part of the standard earthquake coverage. The Cedent's negotiating position depends on the cedent's exposure profile - a cedent with significant Texas or Oklahoma property book wants explicit inclusion; a cedent with only California seismic exposure cares less.

Aftershock aggregation. The 168-hour window covers the main event plus aftershocks within that window. Aftershocks outside the window are a separate occurrence. The 2008 China Sichuan earthquake sequence and the 2011 Tohoku earthquake sequence are reference cases - both produced aftershocks well outside any 168-hour window, and reinsurers treated post-window aftershocks as separate occurrences.

Ultimate Net Loss Definition

Ultimate Net Loss (UNL) is the loss amount the reinsurer cedes against. The UNL definition varies by treaty and is one of the most negotiated clauses. Standard components: paid losses plus outstanding loss reserves plus IBNR (incurred-but-not-reported reserves) less reinsurance recoveries from other treaties less salvage and subrogation recoveries; defense and ALAE (some treaties include in UNL; others exclude); ECO/XPL (Extra Contractual Obligations and Excess of Policy Limits) typically excluded but specific wording matters.

The UNL definition determines what the reinsurer's cession applies against. A wider UNL (including ALAE, including IBNR computed on a specific basis) produces a larger cession base and therefore higher reinsurer payout. A narrower UNL produces a smaller cession base and lower payout. The economic difference can be substantial - for a $500M gross commercial-property loss across a treaty year, a 30% quota share with ALAE-included UNL produces $150M cession; with ALAE-excluded UNL on a 20% ALAE ratio, the cession drops to $125M (the reinsurer keeps $25M of what would have been cession).

2026 negotiation issues. ALAE treatment is the most-negotiated UNL component. The cedent prefers ALAE included in UNL (increases cession base; reinsurer pays a share of the cedent's defense costs and external counsel). The reinsurer prefers ALAE excluded (limits cession base; ALAE volatility stays with the cedent). Salvage and subrogation timing: when does the reinsurer get credit for recoveries - at receipt, at recognition, on an annual reconciliation cadence? ECO/XPL: typically excluded, but Stowers-style exposures (Texas bad-faith) and §624.155 exposures (Florida) are increasingly addressed explicitly. Bermuda-form-style treaties have UNL definitions specific to Bermuda Form 004/006 coverage architecture (covered in the Bermuda Form lesson).

Original Conditions Clause

The Original Conditions clause provides that cessions to the reinsurer are subject to the original conditions of the underlying policy. If the cedent's underlying policy contains a specific exclusion or limit, the cession to the reinsurer is similarly limited. The reinsurer cannot expand the cession beyond what the underlying provides; the cession follows the underlying terms.

This clause is foundational because it aligns the reinsurer's exposure with the cedent's underwriting decisions. The reinsurer is not exposed beyond what the cedent has written; the cedent's discipline in underwriting flows through to the reinsurer's exposure.

2026 critical wording. Reinsurer's wording in 2026 may add representations and warranties about the cedent's underwriting practices, including AI use. Example 2026 Swiss Re wording: "Cedent represents that all underwriting decisions are made in accordance with cedent's documented underwriting guide, and that any AI-assisted underwriting decisions are reviewed by a credentialed underwriter consistent with the NAIC Model Bulletin §4.3 governance principle." The cedent's negotiating position: limit to representations of cedent's standard underwriting practice; preserve flexibility in the AI-governance language; avoid affirmative AI-disclosure obligations that go beyond what state DOIs require. The cedent's general counsel reviews the AI representation language alongside the cedent's CRO and CUO.

Errors and Omissions Plus Sunset Clauses

Errors and Omissions (E&O) clause. Protects the cedent from inadvertent errors in cession - reporting a policy late, misclassifying a risk, omitting a cession from a bordereau - up to a specified materiality threshold. Standard wording: "any error or omission by the cedent in reporting cessions shall not prejudice the reinsurer's obligation, provided the cedent corrects the error promptly upon discovery and the error is not the result of cedent's gross negligence or willful misconduct." The E&O clause is cedent-favorable; reinsurers sometimes negotiate tighter terms requiring formal notification within a specified window (e.g., 30 days) and demonstration of good-faith effort to discover the error.

Sunset clause. Time limit on the cedent's ability to make claims under the treaty after the treaty period expires. Standard: 3-7 years post-expiration. After sunset, the cedent cannot claim against the treaty even if the loss originated during the treaty period - the cession is final and the reinsurer's reserves can be released. The cedent's negotiating position: longer sunset preferred (gives time for slow-developing claims like asbestos, environmental, latent injury); the reinsurer prefers shorter sunset for actuarial certainty and reserve release. The 2026 norm: 5 years for property; 7 years for casualty; longer for long-tail mass-tort (PFAS, opioid, talc).

The interaction with Bermuda Form. For cedents with Bermuda Form 006 high-excess casualty layers (covered in the Bermuda Form lesson), the sunset clause on the underlying treaty must accommodate the Bermuda Form's reporting trigger. A 5-year sunset on the underlying treaty plus a 15-year reporting trigger on the Bermuda Form layer creates a coverage gap if the cedent's notification to the Bermuda carrier happens after the underlying treaty's sunset.

Reinstatement Language

Cat-XOL treaties typically include reinstatement provisions: after a covered loss exhausts the treaty layer, the layer reinstates for additional coverage. Standard: 2x reinstatement (the treaty layer can be exhausted twice in the treaty period). Reinstatement premium: additional premium paid for the reinstatement, usually pro-rata of the original premium based on the remaining treaty period.

The mechanics on a worked example. Heartland's $50M xs $100M cat-XOL layer at 22.8% rate-on-line costs $11.4M annual premium. After one full-layer cat event in June, the layer reinstates for the remainder of the policy year (10 months / 12) at 10/12 of the original premium times the reinstatement rate (often 100% of pro-rata premium for a full reinstatement; some treaties use 100% of full premium regardless of timing). The reinstatement adds $9.5M premium in the example. After a second full-layer cat event in October, the layer reinstates again (the 2x reinstatement) at the remaining period's pro-rata. After a third event would exhaust the 2x reinstatement and leave the cedent unhedged for the remainder of the year.

2026 issues. Reinstatement availability in hard markets: reinsurers may offer only 1x reinstatement or no reinstatement in particularly stressed years (post 2024-2025 cat losses pushed many 2026 renewals to 1x). The cedent must negotiate or accept reduced coverage. Reinstatement pricing reflects expected loss; in hard cat-loss years, reinstatement premium is significantly elevated above the historical 100%-pro-rata benchmark. The cedent's negotiating brief addresses both reinstatement count and reinstatement pricing.

AI-Driven UW Representations - 2026

New in 2026 treaty wordings: reinsurers add representations and warranties about the cedent's AI use. The drivers: reinsurer concern about AI-related cedent losses (model errors producing under-priced cessions; regulatory exposure from biased pricing models; bad-faith exposure from AI-driven claim denials); reinsurer's own treaty actuary's need to validate that cedent's risk-selection discipline holds even with AI augmentation; broader market move toward AI governance documentation.

Example representations. "Cedent represents that AI/ML systems used in underwriting and pricing are subject to documented governance, bias testing per the cedent's home-state DOI requirements (including Colorado Reg 10-1-1 where applicable), and credentialed-reviewer oversight per the NAIC Model Bulletin §4.3 standard." "Cedent will provide the reinsurer with an AI inventory upon request including model purpose, validation methodology, monitoring discipline, retirement criteria, and recent performance metrics." "Material misrepresentation regarding AI use shall be grounds for treaty rescission or non-renewal." "Cedent shall notify the reinsurer of any material AI-system failure, AI-related regulatory action, or AI-related claim that exceeds the cedent's authority threshold."

The cedent's response. Agree to representations consistent with the cedent's actual AI governance posture; negotiate the scope and definition of "AI systems" to avoid overreach (does the chat-bot for FNOL intake count? does the spreadsheet-embedded automation count?); preserve carrier confidentiality of the AI inventory while providing a high-level summary suitable for treaty broker disclosure. The cedent's general counsel reviews the AI representation language; the CUO and CRO sign off; the chief actuary attests to the bias-testing posture.

The 2027 horizon. Reinsurer expectations on AI representations are expected to tighten in 2027 renewals. The cedent that has documented AI governance, regular bias testing, and a defensible AI inventory will navigate 2027 renewals at favorable terms; the cedent without that documentation may face higher rates, stricter representations, or capacity constraints.

Munich Re vs. Swiss Re vs. Hannover Re - Same Risk, Three Wordings

Munich Re wording 2026. Emphasizes the Original Conditions clause and cedent's underwriting discretion preserved; standard 72-hour Named Storm Hours Clause plus 168-hour Earthquake Hours Clause; UNL includes ALAE on property; 2x reinstatement standard; AI representations limited to cedent's documented practice and the bias-testing-per-home-state-requirements language. Sunset clauses 5 years property, 7 years casualty. E&O clause cedent-favorable with standard correction-upon-discovery framing. Munich Re is the broadest-book reinsurer in the global market and tends toward consistency with industry standards.

Swiss Re wording 2026. Aggressive on AI-driven UW representations; tighter E&O clause requiring formal notification within 30 days of discovery plus demonstration of good-faith discovery effort; UNL may exclude ALAE depending on line (cedent-unfavorable for casualty); reinstatement pricing varies by line; audit rights extensive. Swiss Re is positioned at the front edge of treaty-wording evolution and uses Magnum (its AI-assisted underwriting platform) extensively in its own underwriting - which informs its expectations of cedent AI governance.

Hannover Re wording 2026. Emphasizes loss-control reporting cadence and the cedent's cession bordereau accuracy; standard hours clauses (72/168); UNL definition aligned with industry standard; sunset clauses on the longer end (5-7 years on property; 7-10 years on casualty depending on long-tail exposure); AI representations balanced - Hannover Re asks for inventory but is less aggressive on rescission language. Hannover Re's positioning is cedent-friendly on long-tail exposures and rigorous on bordereau discipline.

The cedent's negotiating brief. Compare the wordings clause-by-clause; identify the most favorable wording per clause across the three; build a composite negotiating position; rank the concessions the cedent will trade in negotiation. Example: the cedent wants Munich Re's E&O clause (the most cedent-favorable), Hannover Re's sunset (the longest), and Swiss Re's pricing (if Swiss Re's quote is competitive on rate-on-line) - negotiate a hybrid where the cedent trades concessions on AI representations to Swiss Re in exchange for the longer sunset and the cedent-favorable E&O. AI-assisted markup tools (LexisNexis Practical Guidance, contract-analysis platforms, and treaty-specific tools like the offerings from Send Flow's contract module) parse all three wordings and produce a comparison matrix in 4-8 hours vs. 30-50 hours manual. The treaty broker reviews; the cedent's reinsurance accounting director signs the negotiating brief; the chief actuary and CRO are looped in on material concessions.

Key Takeaways

• Quota-share plus cat-XOL combined treaty architecture. 30% QS on commercial property; $50M xs $100M cat-XOL on Tier-1 wind; coordinated structure ensures gross-loss-after-quota-share triggers cat-XOL attachment correctly.
• Named Storm Hours Clause 72-hour industry standard. Cedent prefers longer; reinsurers may tighten to 60- or 48-hour in hard markets. Compound-event treatment varies - cedent-favorable aggregates flood and tornado within the named storm; reinsurer-favorable separates flood as distinct peril.
• Earthquake Hours Clause 168-hour (7 days) industry standard. 2026 concerns: induced seismicity treatment in Oklahoma/Texas/Kansas wastewater-injection zones; aftershocks within window aggregate, outside window are separate occurrences.
• Ultimate Net Loss definition varies by treaty and is heavily negotiated. 2026 issues: ALAE inclusion/exclusion (most-negotiated component, can swing cession by 15-20%); salvage/subrogation timing; ECO/XPL responsibility (Stowers Texas and §624.155 Florida exposures addressed explicitly).
• Original Conditions clause aligns reinsurer exposure with cedent's underwriting decisions. 2026 reinsurer wordings add AI-related representations that limit cession when AI-assisted underwriting departs from documented governance.
• E&O plus Sunset clauses are cedent-favorable terms. E&O protects against inadvertent errors with prompt-correction requirement; Sunset 3-7 years property, 7+ years casualty. Reinsurer prefers shorter sunset for actuarial certainty and reserve release.
• Reinstatement language 2x standard. Hard markets may reduce to 1x or no reinstatement (2026 saw many renewals tighten to 1x after 2024-2025 cat years). Reinstatement pricing reflects expected loss; cedent's negotiating brief addresses both count and pricing.
• AI-driven UW representations new in 2026. Cedent represents AI governance plus bias testing per Colorado Reg 10-1-1 / NY DFS Circular Letter 2024-7 plus credentialed-reviewer oversight per NAIC Model Bulletin §4.3. Potential treaty rescission grounds for material misrepresentation. 2027 horizon: tighter expectations expected.
• Munich Re vs. Swiss Re vs. Hannover Re - same risk, three wordings. Munich Re consistent with industry; Swiss Re aggressive on AI and E&O; Hannover Re cedent-friendly on long-tail with rigorous bordereau. AI-assisted markup compresses 30-50 hours manual to 4-8 hours; treaty broker reviews; reinsurance accounting director signs the negotiating brief; chief actuary and CRO loop in on material concessions.