RESEARCH DOSSIER
Catastrophe Bond Settlement and Principal Loss Allocation
Abstract. Catastrophe bond settlement is the contract-governed conversion of an eligible catastrophe observation or insured-loss determination into an amount of risk-transfer payment, a corresponding reduction in investor principal, and a legally final reconciliation of remaining rights. This monograph examines the distinct stages of occurrence, notification, measurement, certification, principal allocation, collateral realization, redemption, and dispute resolution. It combines transaction-level analysis of the International Bank for Reconstruction and Development's 2024 Jamaica capital-at-risk notes with comparative evidence on the 2019 Peru earthquake payout and the 2025 Hurricane Melissa payout, the economics of indemnity and non-indemnity triggers, the legal status of collateralized special-purpose arrangements, and accounting distinctions. It derives bounded principal-allocation identities and verifies hypothetical examples for occurrence and aggregate structures, successive event reports, pro rata holders, multi-tranche layers, and maturity extensions. The central finding is that an event's physical severity, an insurance loss, a contractually calculated payout, investor principal impairment, cash settlement, and the investor's market mark are related but distinct variables whose equality must be established by the specific transaction terms. Timing provisions create real exposure to data revision, liquidity restrictions, opportunity cost, and imperfect information. Explicit calculation-agent instructions, observable cutoffs, reserve-release conditions, and reconciliation controls determine when provisional loss expectations become enforceable reductions. The study provides comparative contractual interpretation and original analytic synthesis, without proposing a new proprietary theory or claiming empirical estimation from confidential transaction records.
Keywords: Catastrophe bond; insurance-linked securities; settlement; principal reduction; catastrophe trigger; event calculation agent; indemnity; parametric trigger; industry loss; loss creep; trapped collateral; maturity extension; redemption; basis risk; special purpose vehicle; sovereign disaster risk financing; investor allocation.
Introduction
A catastrophe bond is a contingent claim on capital that investors place at risk against a specifically identified set of catastrophe-related outcomes. Its apparently simple economic exchange—periodic compensation for assuming the chance of a principal loss—rests on a more elaborate sequence of contractual determinations. A hurricane may occur in minutes, the market may price a potential impairment within hours, and an insurer's underlying claims may develop for years. Which of these observations gives rise to a legal claim against the bond depends on the definition of the covered event, the risk period, the trigger measurement, the reporting timetable, the payment waterfall, and the finality clauses.
The World Bank's capital-at-risk notes program describes both special-purpose-vehicle structures and direct IBRD issuances. The latter matter for comparative analysis: in an IBRD catastrophe bond the Bank itself issues notes and enters into a separate risk-transfer contract with the beneficiary, rather than routing the notes through a conventional special-purpose reinsurer. The economic transfer of risk remains intelligible across both structures, although legal obligations, governing instruments, and the identity of the party holding collateral differ. The OECD analysis of transaction engineering describes the ordinary trust-funded SPV design and the consequences of choosing a parametric, industry-loss, modeled-loss, or indemnity trigger.
For this dossier, 'settlement' has two separate senses. Issuance settlement is the closing date on which investors pay the subscription price and the notes are issued; event settlement is the later contractual process that determines risk-transfer claims and applies principal reductions or repayments. The Jamaica 2024 issuance announcement records May 2, 2024 as issuance settlement and December 29, 2027 as scheduled maturity, whereas the 2025 catastrophe produced an early loss-triggered redemption. Conflating issuance settlement with event settlement would invert the direction of cash flows.
The analytical object consists of five questions. What makes an event eligible? What evidence is contractually admissible? How does the loss metric become a payout amount? How is that amount allocated across principal interests and time? When are the remaining claims finally discharged? The answer to each question is transaction-specific. This monograph therefore develops general arithmetic only under clearly stated assumptions, then tests the interpretation against an actual term sheet and official outcomes.
The research method is documentary and comparative. Primary sources include the IBRD 2024 prospectus supplement and terms, official World Bank issuer releases and case studies, official governmental accounts of sovereign risk financing, and European regulatory texts. Supporting scholarship includes Cummins (2008), Cummins and Weiss (2009), Braun (2016), the Federal Reserve Bank of Chicago's 2018 primer, and the OECD's 2024 institutional comparisons. Direct contract language is treated as authoritative for the named transaction; broader market observations are interpreted at the source's population and date. Mathematical illustrations are hypothetical, denominated in U.S. dollars unless explicitly stated, and independent of actual private investor registers.
Definitions, Source Record, and Scientific Boundary
A distinction among event, attachment, claim, and debt reduction
A covered event is a natural or other catastrophe that satisfies the specified temporal, spatial, peril, and data-source criteria. A triggering measurement is the number or defined condition derived from an eligible event: a central pressure value, an estimated industry loss, a modeled portfolio loss, or an ultimate indemnity loss. Attachment is the point beyond which the relevant layer begins to pay. Exhaustion is the point at which the applicable layer has paid its maximum contractual amount. A payout is the amount due under the risk-transfer formula. Principal impairment is a reduction of investors' contractual repayment entitlement, which may fund or economically mirror the payout. Cash transfer is the movement of funds under the relevant settlement instructions. These quantities may have different dates and different measurement rules.
The Chicago Fed's trigger comparison identifies the economically important information distinction: indemnity contracts refer to the sponsor's realized insurance losses, industry-loss contracts to third-party sector estimates, and parametric contracts to a measured physical event. A given earthquake does not itself determine a bondholder's loss; the contract must map it into the specified payout and principal-reduction functions. The NAIC's insurance-linked-securities overview likewise distinguishes insurance-risk trigger losses from collateral credit events. A reduction in a security's secondary-market price because of uncertainty must not be reported as a finalized contractual principal reduction.
It is useful to represent the scientific object by an ordered set of observations and rules: event E, admissible measurements X, trigger function g, contractual payout P, allocation rule A, payment and redemption dates T, and a finality rule F. This notation is an analytical description of existing contract mechanics, not a novel theory. A bond with exactly the same peril E may yield a different legal payout because it uses a different g, a different measurement cutoff, or a different limit. The study therefore refuses the misleading shortcut of treating catastrophe severity as an investor-loss percentage.
Transaction architecture and legal incidence
In a conventional collateralized catastrophe-bond structure, investors subscribe to notes issued by a bankruptcy-remote special-purpose insurer or vehicle. The transaction vehicle enters into reinsurance or a derivative with the sponsor or an intermediary, holds the funded proceeds in qualifying collateral, and pays periodic note coupons. When the specified contract triggers, collateral is applied to obligations owed to the sponsor, reducing funds available to redeem investor principal. The OECD account of typical CAT bond cash flows provides the basic economics, including the separate floating investment return and catastrophe spread.
The exact ranking of sponsor, trust, trustee, issuer, and noteholder rights comes from the indenture, trust deed, retrocession or reinsurance agreement, collateral arrangements, and any derivatives. Some structures employ more than one note class with different layers. A sponsor's protection is secured in economic terms by funded capital, but absolute immunity from operational, legal, investment, or counterparty risk cannot be presumed. Cash realization, custody conditions, legal enforceability, and allowed collateral substitutions all require examination. The historical collateral experience summarized in the NAIC ILS material is a reminder that principal impairment can arise from collateral mechanisms as well as the catastrophe trigger, with distinct causes and attribution.
The IBRD architecture is materially different. The World Bank program description states that the Bank issues the notes, manages proceeds, and makes catastrophe-contingent payments under a contract with the beneficiary. It follows that an SPV trust-waterfall model should not be imposed on the 2024 Jamaica note's legal form. The bond's actual note conditions and risk-transfer agreement determine its legal liabilities. Principal reductions in those notes are an expressly documented feature of the IBRD terms, not an inferred transfer from a hypothetical trust.
Risk layers and basis risk
For an ordinary occurrence-based excess-of-loss layer, let L be the eligible monetary loss, a the attachment amount, and m the funded limit. A simplifying payout function is P(L)=min[m,max(0,L−a)]. Exhaustion occurs at L=a+m. This identity is useful for layer arithmetic, but it assumes a continuous dollar-loss trigger without separate reinstatement, franchise, co-participation, aggregation, time limits, or exclusions. A parametric contract substitutes a contract-specific function of hazard measurements, not the sponsor's economic loss. The distinction is central to sponsor basis risk: an event can generate severe losses with no payout if the index fails its threshold; another event can produce a payment above the sponsor's contemporaneously measured loss.
Cummins and Weiss's insurance–capital-markets convergence study locates catastrophe securities within the broader financing of insurable risks, while the OECD discussion of basis risk explains why non-indemnity triggers can speed sovereign liquidity yet create a gap between damage and recovery. Trigger simplicity is therefore a trade involving observability, verification cost, settlement latency, manipulation exposure, and the economic fit of protection to the sponsor's loss profile.
Sources and evidentiary precedence
A transaction's executed legal documents control specific payment rights; contemporaneous prospectuses explain disclosed terms; calculation-agent event reports apply those terms; issuer notices and official accounts document subsequent implementation; statistical and academic studies support comparisons across classes of instruments. A sponsor press release can establish what the sponsor reported but does not, by itself, audit an insurer's ultimate claim liabilities. This source hierarchy is a practical evidentiary discipline. It does not mean that contracts are immune to applicable mandatory law or that any press release establishes the outcome of a particular private dispute.
The public record is unusually strong for Jamaica's 2024 CAR 136 because it includes the full listed note documentation, the initial transaction announcement, a later World Bank transaction case study, and records of the 2025 catastrophe payout and 2026 successor. They permit close testing of settlement mechanics without assuming access to confidential underwriting files. For private indemnity bonds, unavailable loss development triangles and internal adjustment records constrain any claim about exactly when a particular event finally settles.
Event Notification and Settlement Chronology
The contractual calendar
The settlement chain begins before the catastrophe, when the terms define the risk period, covered geographic area, triggering peril, designated data providers, calculation agent, calculation dates, permissible notices, and payment dates. The observed catastrophe produces a potential event, followed by a notice under the contract, data collection, calculations, a report, any correction process, and principal adjustment. A note can lose principal on a specified payment date that occurs after the event and after the data cutoff. Physical occurrence, legal eligibility, formal notice, report publication, principal reduction, sponsor receipt, and terminal redemption should therefore be recorded as separate fields in a transaction chronology.
The Jamaica 2024 note terms supply a concrete example. An event report is required to determine whether the specified named-storm event occurred, its occurrence date, the calculation of its payout amount, the applicable principal reduction, and the resulting outstanding nominal amount. The document provides that principal reductions are scheduled at specified interest payment dates and the redemption date, with an event report cutoff five business days before the relevant date. That is an operational timing rule, not a general law applying to every catastrophe bond.
Suppose, as a hypothetical timing example, a $100 million note has a monthly reduction date on a Friday; the required event report is delivered after the contract's five-business-day cutoff. Even when the result is clear to market participants, the principal reduction cannot be assumed to occur on that immediately upcoming date. Under a rule like the Jamaica structure it falls to a later eligible payment date. The economic exposure may already be priced, but the contractual outstanding nominal amount remains governed by the document until the effective principal-reduction date. Any exact business-day calculation must account for the applicable holiday calendars and notice provisions.
Measurement and reporting under parametric triggers
A parametric payout requires a definition of the observable hazard and an agreed data hierarchy. For tropical cyclones, this may include location of the storm center, pressure or wind, reporting source, admissible time window, and calculation of multiple geographic cells. The agent's job is to implement the formula, not to estimate the sponsor's final reconstruction bill. A contract may freeze event parameters at a stated observation date even when an upstream meteorological agency subsequently revises a storm track. Such a freeze reduces open-ended data revision while making the exact version of the record consequential.
The Jamaica 2024 catastrophe-linked terms specify a calculated central-pressure-based payout for covered cells. Where a cell qualifies between two defined pressure thresholds, the documentation uses an interpolating percentage that rises as calculated central pressure falls; sufficiently low pressure produces full payout for the cell, subject to contract aggregation and the overall cap. The public World Bank 2024 case study characterizes the payout as a stepwise linear sliding scale, with 30%, 70%, and 100% principal-payout reference rates. The authoritative formula, including a particular cell's thresholds, remains the detailed note documentation and associated defined data.
A named storm can cross multiple geographic cells or continue beyond an initial calculation window. The Jamaica terms allow subsequent event reports and specify that incremental amounts are calculated with reference to the difference between a later and an earlier payout rate, with a zero floor when the later measure does not increase the payout. This guards against charging the same investor principal twice for a revised measurement of the same event. It also illustrates why 'report count' is an invalid substitute for 'distinct catastrophe loss count.'
Indemnity, industry-index, and modeled-loss paths
Under indemnity settlement the sponsor's relevant ultimate net loss must be measured under the contract. Claims are reported, reserved, adjusted, litigated, recovered from salvage or subrogation, and allocated to the eligible account. Reinsurance recoveries, occurrence definitions, aggregate deductibles, and hours clauses can affect the reported layer loss. If the loss remains unsettled, a note that would otherwise mature may need to preserve collateral while the exposure develops. The Federal Reserve Bank of Chicago's historical comparison reports materially slower payout for indemnity structures than for index or parametric structures within the historical context studied; its historical averages should not be converted into universal deadlines.
Industry-loss index bonds substitute a third-party estimate of the industry's insured catastrophe losses, sometimes weighted by geography and peril. The sponsor avoids dependence on its own claims-handling completion but accepts the risk that an industry estimate differs from its experience. The OECD trigger mechanism survey identifies Property Claims Services in the United States and PERILS in Europe as relevant industry-loss information channels. Updates to an index may alter an anticipated payout until the contract's final observation or settlement date. Modeled-loss structures use defined hazard inputs and a predetermined portfolio or model protocol; transparency depends on the completeness and stability of those specifications, including model versions and assumptions.
The chronology is therefore an information-production process. Parametric bonds can reach a measurement-based decision quickly if the data and methodology are available, but disputed data, infrastructure failure, or exceptional event complexity may still delay a report. Indemnity bonds can have a close statistical association with the sponsor's true loss, yet their ultimate-loss measurement tends to require more claim development. Speed is obtained through the choice of a settlement variable and verification architecture, not merely by labeling an instrument 'parametric.'
Official catastrophe experience: 2017, 2019, and 2025
Mexico's 2017 earthquake experience is a widely documented example of full principal-linked catastrophe recovery for an eligible parametric class. The OECD country transaction account describes the magnitude-8.1 Chiapas earthquake and a $150 million Class A payout. The evidence concerns the defined class, not a conclusion that every class of Mexico's then-outstanding multi-peril program was exhausted. Mexico's claim cannot be equated with the country's total disaster cost.
Peru demonstrates a distinct partial outcome. The World Bank's Peru results account reports that a $200 million earthquake catastrophe bond participated in through the Pacific Alliance transaction paid $60 million after the 2019 magnitude-8.0 earthquake. The arithmetic is 60/200=30% of the referenced $200 million protection. This published outcome illustrates proportional contractual loss allocation; it does not identify confidential beneficial-owner positions or prove that all uncovered national losses were limited to $60 million.
Jamaica provides a recent complete event-to-redemption case. The original World Bank Jamaica 2024 transaction notice specifies a $150 million parametric note and scheduled maturity in December 2027. Hurricane Melissa struck Jamaica on October 28, 2025. The November 2025 World Bank determination records that the independent calculation agent found the contract's path and pressure conditions satisfied for a full $150 million payout. The later World Bank payout and redemption record gives December 1, 2025 as Jamaica's receipt date and reports that the notes were fully redeemed. The latter record establishes the difference between a November event determination and December cash receipt.
The observed gap between October 28 and December 1 is 34 calendar days. This derived duration includes all intervening stages and must not be presented as the calculation agent's processing time, which the 2026 World Bank successor case study describes separately as an approximately nine-day confirmation after landfall. On May 26, 2026 the World Bank settled a new $200 million Jamaica note; the Government of Jamaica's replacement announcement identifies the prior $150 million coverage as fully paid in 2025. The two bonds are different issues; the new issuance does not undo the prior bond's realized loss.
Settlement timing and economic consequences
The duration between hazard observation and final principal reconciliation changes liquidity, coupon expectations, duration risk, market-making spreads, and capital available for reinvestment. If a fund owns a note at par before an event and estimates a 60% loss, the interim quotation can reflect that expectation before any legal reduction. Accounting measurement and reported net asset value may respond on an earlier date than contractual principal. The final cash flow is still governed by the note and the effective settlement chronology. The resulting timing gap is economically substantive even if every eventual loss amount is correctly measured.
A quantitative timing metric should record at least the event timestamp, first notice, publication of a preliminary estimate, definitive report, effective principal reduction, and final cash payment. These events support measurement of elapsed times, but available market studies rarely disclose identical timestamps for all bonds. Comparing only the interval from a hurricane to a press release for one transaction with the interval to complete claims runoff for another would exaggerate structural differences. Appropriate comparison fixes the endpoint and reports residual censoring where an instrument remains outstanding or extended.
Contractual Development Periods and Extensions
Why an announced maturity is conditional
A maturity date is a legal promise qualified by its terms, including relevant catastrophe-related extension rights. In an untriggered security the expected principal may become payable on the scheduled date. An eligible catastrophe shortly before that date can make the ultimate amount uncertain while the calculation agent, insurer, or index provider completes the relevant observation or loss development. A well-drafted extension clause preserves the contractual risk-transfer period and prevents premature distribution of collateral that may be needed for an existing covered event. Its exercise also changes the bondholder's liquidity and return profile. The extension decision requires separate analysis of event coverage, outstanding principal, term-specific notice conditions, and the date when the sponsor's claim becomes quantifiable.
The 2024 IBRD Jamaica note conditions provide a particularly precise example. The scheduled maturity was December 29, 2027. Under defined notification conditions, maturity could extend initially to January 29, 2028 and then automatically by further monthly periods, subject to a contractual maximum of April 29, 2028. The text refers to an extension notice delivered by the Government of Jamaica and also covers a timely potential-event notice for which the required event report has not arrived by the specified pre-maturity cutoff. Different business-day advance-notice requirements apply to these alternatives. The particular details should be read from the 2024 document rather than generalized as a four-month market standard.
An important economic detail is that the Jamaica 2024 contract also distinguishes the ordinary catastrophe-risk-margin period from an extension period. The prospectus supplement's coupon provisions provide a 7.00% annual risk margin during the risk period but a 0.10% annual risk margin during an extension period, alongside other coupon components and conditions. Investors therefore cannot forecast extension compensation simply by extending the full catastrophe-risk spread for an arbitrary number of months. The applicable calculation depends on the actual note, timing, outstanding nominal amount, and interest convention. A comparison that treats the stated annual risk margin as a fixed coupon throughout the note's remaining life would misstate the economics.
Development of indemnity losses
Indemnity losses evolve as underlying claims are reported and adjusted. An initial estimate can move because of delayed notification, litigation, repair-cost inflation, demand surge, policy interpretation, subrogation, salvage, or catastrophe clustering. A reported insured industry loss may also be revised as insurers submit updated information. The expression 'loss creep' denotes such post-event upward development, but not every later change constitutes a new catastrophe, and some revisions reduce loss estimates. An extension arrangement can hold capital against the relevant potential payout while the cedent and agents resolve these developments. There is no universal rule that a specific percentage of collateral should remain trapped merely because a sponsor has issued an early loss estimate.
Consider a hypothetical $100 million indemnity note providing per-occurrence protection above a $500 million sponsor net-loss attachment, exhausted at $600 million, with no reinstatement. At the initial estimate of $530 million, the expected payout is $30 million. If an updated eligible loss reaches $580 million, the expected payout becomes $80 million. If final adjusted eligible loss reaches $560 million, the contractual payout is $60 million. The changes are $30 million → $80 million → $60 million. They reflect revisions to the same underlying eligible event rather than three separate principal-depletion events. If the contract calls for provisional releases and loss adjustments, the amounts and their legal finality must be computed under those clauses, not by treating the largest forecast as permanently paid. Under a contract permitting only final settlement, the early numbers are valuation estimates, with the final reduction determined later.
The Chicago Fed's comparison of trigger timelines supports the institutional mechanism: actual insurer loss verification generally creates longer claims-development paths than directly observable physical metrics. This does not imply that all parametric instruments mature without extensions, because particular data-observation windows or disputes can keep even a simple physical trigger unresolved. An agent may require a corrected source file or perform additional calculations after a qualifying event near expiry. The governing distinction is the contractually required information still missing on the release date.
Trapped collateral, reserves, and the economic cost of delayed release
Trapped collateral refers to funded assets retained because the transaction still faces a possible eligible obligation. Its existence and amount are not the same as a final insured loss or final investor impairment. A $100 million tranche might retain $100 million of collateral against a plausible maximum loss while its current expected impairment is $20 million; the retained balance is a liquidity constraint, not a $100 million realized loss. Alternatively, partial release of $70 million could leave $30 million held against unresolved claims. Both examples require the actual contract to allow the stated release. Investor claims to coupons, investment returns, and residual capital depend on the relevant arrangements.
The OECD assessment of catastrophe-bond constraints identifies illiquidity and uncertainty in market valuation among investor concerns. A practical internal monitoring report should therefore distinguish: original funded principal; legally reduced principal; projected impairment; principal subject to release restriction; realizable trust value; expected final redemption; and outstanding permitted claims. Adding projected impairment and trapped collateral as though they were independent losses can double-count the same potential catastrophe exposure.
For illustration, suppose $40 million of otherwise distributable principal is delayed for six months, and the investor's marginal opportunity return on comparable liquid risk is 5% annually. A simple non-compounded opportunity-cost approximation is $40 million × 0.05 × 0.5 = $1 million. This is not a contractual claim for damages, a risk-free valuation, or a substitute for discounting actual cash flows. It demonstrates why a potentially fully recoverable balance still carries economic cost if its release date is uncertain. Differences between actual extension coupons and replacement-investment opportunities can increase or reduce the investor's net cost; the sign cannot be inferred from the amount held alone.
Reserve release as a controlled information decision
Reserve release should be tied to objective conditions: completed reporting periods, updated accepted loss estimates, exhaustion of notice windows, calculation-agent sign-off, termination of pending disputes, and maximum contractual liabilities. A process might calculate a cautious retained amount based on the upper bound of the unpaid eligible obligation and release the remainder. The formula for a specific security, however, could contain floors, full-extension triggers, priority payments, partial-extension mechanics, or provisions that prevent release until all qualifying events are conclusively addressed.
From the investor's perspective, reducing a held reserve increases deployable capital and may lower secondary-market uncertainty. From the sponsor's perspective, early release creates the risk that an ultimately valid claim lacks sufficient funding. The contract distributes these consequences ex ante by specifying how much discretion exists, who exercises it, and what evidence is binding. Well-defined revision cutoffs matter as much as the initial risk model. A model that precisely forecasts annual expected loss but ignores the duration and uncertainty of settlement reserves cannot fully describe the investor's realized capital availability.
Expiry, eligible event, and post-risk-period development
The end of an insurance risk period is not necessarily the end of the bond's exposure to payment on events occurring before that date. The IBRD note's covered-event and extension language expressly distinguishes the timing of a named storm's occurrence from its subsequent continuation or reporting, subject to the stated event definition. A valid risk-period event may lead to a later reduction. Conversely, a storm arising outside the agreed risk period is not transformed into an eligible event by a later payment date or by a sponsor's continuing need for relief. In an indemnity case, the exposure may remain contingent on the adjustment of claims from pre-expiry eligible occurrences while new post-expiry occurrences are outside coverage.
This separation is essential in actuarial modeling. A catastrophe simulation ordinarily produces annual or event losses. A settlement simulation must additionally produce event reporting dates, delay distributions, revised estimates, contractual cutoffs, extension elections, and ultimate payment dates. It can then estimate not only an expected principal loss but the distribution of time-weighted capital restrictions. Those calculations require data distinct from an event-loss distribution. Treating all covered-event payments as if they occur instantly at the hazard timestamp overstates liquidity and obscures maturity-extension risk.
Partial and Total Impairment Allocation
Baseline principal accounting identity
Let N0 be the original aggregate nominal amount of a single note class. Let Dk ≥ 0 denote the legally effective catastrophe-related principal reduction on settlement date k and Rk ≥ 0 an actual partial repayment of outstanding principal on that date. Assuming no write-ups, no new issuance into the class, and no additional fee deductions from nominal balance, the outstanding nominal amount after K dates is
At each date, the reduction must be capped by the remaining outstanding balance. A convenient operational expression is Dk = min[max(0,Ck − Ck−1), Nk−], where Ck is the cumulative recognized event payout to the contractual cutoff and Nk− is the nominal amount immediately before applying the new reduction. The validity of this expression depends on whether the applicable agreement uses cumulative event reports, individual-event increments, distinct classes, or other offsets. It reproduces the broad logic of the 2024 Jamaica principal-reduction clause, which compares cumulative reported payouts at successive cutoffs and imposes a cap at outstanding nominal.
If a class has $150 million initial nominal and recognizes a single $150 million principal reduction with no prior partial repayment, the end-of-period nominal is zero. If it recognizes $45 million, the remaining nominal is $105 million. If it has already repaid $30 million of nominal before a prospective catastrophe adjustment, the remaining $120 million is the maximum possible subsequent principal reduction. A payment of interest is not itself a repayment of nominal. Bond-market total return combines coupons, principal recovery, and purchase price; principal-loss percentages concern the legal principal denominator specified by the contract.
Reconciliation of cumulative and incremental reports
The cumulative-reporting approach avoids double counting when one event is revisited. Take a hypothetical $150 million nominal note with the following recognized reports and dates. No partial repayments occur. The table treats the cumulative amount as already validated by the agent and by all contract eligibility conditions.
| Effective date | Cumulative recognized payout | New principal reduction | Outstanding nominal |
|---|---|---|---|
| At issue | $0 million | $0 million | $150 million |
| First eligible date | $30 million | $30 million | $120 million |
| Second eligible date | $75 million | $45 million | $75 million |
| Third eligible date | $75 million | $0 million | $75 million |
| Fourth eligible date | $180 million | $75 million | $0 million |
At the fourth date the difference in cumulative reports is $105 million, but the remaining legal nominal is $75 million; the reduction is capped at $75 million. The $180 million figure deliberately exceeds bond capacity to test the cap and is an abstract series, not a real transaction's payout. If instead a contract never permits cumulative event payouts above $150 million, an agent would cap the reported payout at the source rather than present $180 million. The result is the same nominal floor but the document's reporting method differs. The ledger's essential controls are monotone effective reductions, an explicit partial-repayment register, and a zero minimum nominal.
The Jamaica 2024 report provisions further define how subsequent event reports calculate increments. In particular, a subsequent report for the same named storm determines an additional amount only to the extent its relevant payout rate rises above the original measurement. An interim market estimate that declines is not mechanically a reversal of a legally effective prior principal reduction. Whether any reversal or compensation mechanism is available is a separate contract-law question, and should not be inserted into a note balance equation without express terms.
Pro rata allocation to holders
When the note conditions require pro rata allocation among holders within one class, the legal loss applies proportionally to each holder's outstanding nominal position, not according to purchase price or the holder's internal investment valuation. If class nominal before an effective reduction is N and holder i owns qi nominal, its share of a D principal reduction is di=D(qi/N), assuming positions are evaluated on the contractual record date and no additional holder-specific adjustments exist. The sum of holder reductions equals D because the ratios qi/N sum to one.
For a $150 million class with holders owning 50%, 30%, and 20%, a $45 million principal reduction produces holder impairments of $22.5 million, $13.5 million, and $9 million, leaving $52.5 million, $31.5 million, and $21 million nominal. The result is independent of whether one investor acquired its position below par before the catastrophe. Investors who transact between an early market warning and the reduction date transfer economic exposure under trade terms and settlement conventions; the issuer continues to apply reductions to the legally recorded beneficial interests at the prescribed time.
The IBRD CAR 136 pro rata term explicitly supports within-class proportional principal reduction. It does not establish that different catastrophe-bond classes across an issuance series must share losses pro rata with each other. Separate classes may have different triggers, risk periods, attachment points, currency, seniority, and payout caps. The relevant unit of calculation must be identified before any attribution of investor losses.
Loss layers and different classes
A hypothetical two-layer catastrophe bond program demonstrates how a sponsor's event loss can exhaust one class without impairing another. Suppose the sponsor has an eligible net occurrence loss L; Class A provides $100 million of protection attaching at $400 million, while Class B provides $150 million attaching at $500 million. Ignore reinstatements, co-participation, and other contract conditions. The simple class payout functions are PA(L)=min(100,max(0,L−400)) and PB(L)=min(150,max(0,L−500)), with every amount expressed in millions. Class A is first affected at loss above $400 million and exhausted at $500 million; Class B begins to respond above $500 million and exhausts at $650 million.
| Hypothetical eligible loss | Class A payout | Class B payout | Total sponsor payout |
|---|---|---|---|
| $350 million | $0 million | $0 million | $0 million |
| $450 million | $50 million | $0 million | $50 million |
| $550 million | $100 million | $50 million | $150 million |
| $700 million | $100 million | $150 million | $250 million |
These figures represent layer payoffs to a hypothetical sponsor and possible principal depletion for fully funded one-loss class structures. They are not a description of Jamaica's CAR 136, whose payout is parametric. A first-loss class can be fully written down while a higher-attaching class remains intact. A senior/subordinated credit-waterfall structure may implement a different ordering; the exact legal ranking must prevail over the pedagogical attachment-layer picture.
Per-occurrence and aggregate protections
An occurrence bond treats each eligible occurrence under its stated occurrence definition and per-event limit. An aggregate bond accumulates specified eligible losses during an agreed period before applying an annual or term aggregate attachment and exhaustion. Mathematical shorthand can conceal significant contractual differences, including deductible erosion, hours clauses, qualifying-event thresholds, exclusions, and event counting.
Suppose an aggregate contract attaches at $200 million of eligible cumulative period loss and covers the next $100 million. Hypothetical losses of $80 million, $100 million, and $90 million accumulate to $270 million, producing $70 million payable at the final aggregate calculation, subject to contract reporting rules. If the first two observations are below the attachment, it does not follow that they are irrelevant; together they erode $180 million of the $200 million threshold. Under a per-occurrence structure with a $200 million attachment for each single event, none of those separate events would generate payment. The apparent 'same deductible' creates wholly different results because the unit of aggregation differs.
Catastrophe settlement records should identify the event number and measurement boundary, calculation date, aggregate deductible exhausted to date, additional payable amount, prior paid amount, and remaining covered limit. Failure to align these fields can produce artificial overstatement when one catastrophe appears in multiple estimate vintages or artificial understatement when a late reported loss legitimately contributes to the aggregate threshold.
Principal loss versus investor total return
A 100% principal loss is a contractual statement about original or outstanding nominal, not necessarily a 100% loss on an investor's entire holding-period return. Coupons received before a triggering event may offset part of the investment's cumulative cash outflow; an investor purchasing at a discount may have a different realized dollar loss from an investor subscribing at par. Conversely, a partial principal write-down may be accompanied by a reduced future coupon base and delayed or restricted liquidity. Accurate attribution distinguishes ex-ante expected loss, mark-to-market movement, realized principal adjustment, coupon receipt, expenses, and realized proceeds from a secondary-market trade.
Braun's empirical primary-market pricing research identifies expected loss as the principal empirical determinant of issue spreads across a historical international sample, alongside territory, sponsor, reinsurance cycle, and comparable corporate bond spreads. The implication for settlement is indirect but important: an expected-loss premium priced at issuance is compensation for contingent risk, while ex-post payout amounts derive from contract and event evidence. A high spread cannot establish the precise amount lost after any particular hurricane. Nor does a bond's market discount automatically constitute an enforceable payable amount to the sponsor.
Principal reconciliation under partial repayment
Partial repayment must be distinguished from principal reduction: repayment gives investors cash and reduces outstanding nominal, while impairment extinguishes nominal without paying that portion to the investor. Consider a hypothetical $100 million note that first repays $20 million and later suffers a catastrophe principal reduction of $50 million. Its remaining legal nominal is $30 million; investors have received $20 million principal back, lost $50 million principal, and still own a $30 million claim, plus any contractual coupons. If the event occurred before the $20 million repayment and coverage or calculation rules use original nominal, the final economics may differ. The order of operations is therefore prescribed rather than assumed.
The IBRD 2024 redemption terms express redemption by reference to outstanding nominal after applicable principal reductions and partial repayments. This is a stronger accounting anchor than a secondary-market quotation or sponsor headline. It permits reconciliation of noteholder rights even when multiple payment dates, partial settlements, or event reports intervene.
Maturity, Dispute, and Final Reconciliation
Legal finality and operational finality
Final reconciliation determines which observations count, which calculations control, whether timely objections remain open, what principal survives, what cash is payable, and whether contingent contract obligations have terminated. 'Final' can refer to the event calculation, the principal debit in note records, the payout to a sponsor, or the discharge of a note. The event calculation may become binding before cash reaches the beneficiary, as the Jamaica 2025 chronology demonstrates. Finality must therefore be indexed to the named legal obligation rather than assigned as a single undifferentiated date.
For a traditional special-purpose insurer structure, the final process also involves trust-account instructions, liquidation or transfer of admissible collateral, settlement of permitted administrative expenses, payment of any reinsurance obligation, and distribution of residual principal under the agreed ranking. In an IBRD note, the redemption amount and its payment arrangements derive directly from the issuer's note terms, with the separate sovereign contract explaining the beneficiary's rights. Each transaction requires a closing balance sheet and evidentiary trail that reconcile the economic transfer with legal extinguishment of claims.
Correction of calculation-agent errors
An event calculation agent typically applies prescribed methodologies to data and publishes a report. That technical determination can have significant financial consequences. A finality clause may make its calculations binding, subject to a narrow exception for manifest error, rather than opening every modeling judgment to unlimited retrospective relitigation. The precise notice window and remedy are contractual choices.
The CAR 136 manifest-error procedure provides a documented example. Subject to the note's publication mechanics, a potential manifest error must be identified in a written notice received within three business days after the event report is first made available on the designated site. IBRD then determines, in consultation with the event calculation agent, whether the potential issue is a manifest error, under its stated time condition. An affected event report does not produce the relevant reduction until an amended and restated report cures the error. This is more exact than saying holders have an unlimited right to contest any modeled result. The clause does not justify assuming that a different issuer uses the same challenge mechanism.
An effective review should preserve the original report, the version of event data used, notification time, manifest-error notice, adjudication outcome, corrected report, and revised principal ledger. A later publicly updated storm record is not automatically sufficient to reopen a closed event if the contract fixes data as of an earlier date. A reporting error, a changed scientific understanding of the event, and disagreement over the formula's risk design are different disputes, with potentially different remedies.
Governance of sources and calculations
The calculation process should be auditable at a transaction-specific level. Critical controls include separation of sponsor and agent responsibilities; unambiguous data identifiers; reproducibility of geographic or index calculations; preservation of formula versions and timestamped inputs; documented rounding and currencies; confirmation of the applicable class and nominal denominator; independent reconciliation of nominal balances; and a recorded escalation pathway for exceptions. The official Jamaica documents identify AIR Worldwide Corporation as the relevant modeling and calculation provider for the 2024 issuance. The World Bank's retrospective case study identifies the issuer's service-provider procurement and the observed outcome.
In index and parametric transactions, the quality of data-provider governance is part of the value of the instrument. Physical parameters that are openly observable can reduce information asymmetry, but contractual selection of stations, grid definitions, source revisions, and reporting deadlines can still favor one outcome over another when a storm sits near a threshold. The OECD's measurement-infrastructure analysis emphasizes independent data, robust infrastructure, and testing for basis risk. These measures improve reliability without guaranteeing that every sponsor loss is covered.
Jurisdiction, regulatory scope, and credit protection
Legal classification differs across jurisdictions. In the European Union, the Solvency II definition of an SPV refers to an undertaking that assumes insurance or reinsurance risk and fully funds that exposure through debt or other financing subordinated to its reinsurance obligations. Solvency II Article 211 requires supervisory approval for the establishment of such vehicles within the relevant national framework, while Delegated Regulation Article 326 specifies conditions relevant to fully funded status, including assets at least equal to aggregate maximum risk exposure and ability to pay liabilities as they fall due. These provisions bear on European qualifying SPV arrangements; they are not themselves the governing note law of an IBRD-issued Jamaica security.
Supervisory recognition of risk mitigation and investors' contractual rights are separate questions. A reinsured entity may have to demonstrate legal enforceability, effectiveness, and sufficient counterparty or collateral quality to recognize protection in prudential calculations. That recognition does not fix the final catastrophe loss. Conversely, an investor can suffer a contractually valid principal reduction even when the sponsor's accounting treatment of the recovery is disputed or deferred. A serious reconciliation file records the governing jurisdiction of the issuer and note, the relevant reinsurance or derivative contract, supervisory authorizations where applicable, collateral custody, currency, sanctions and payment-system constraints, and available enforcement processes.
Accounting and valuation boundaries
Financial reporting follows a measurement architecture that may differ from the legal principal ledger. For an investor, a catastrophe-linked note is a financial instrument whose classification and measurement depend on the entity's accounting basis, business model, and contractual cash-flow characteristics. Under IFRS 9 Financial Instruments, the contractual-cash-flow and business-model conditions affect classification; a catastrophe-contingent principal feature requires instrument-specific assessment. This monograph makes no blanket classification of all cat bonds as amortized cost, FVOCI, or fair value through profit or loss. Mark-to-market valuation may recognize deterioration in expected recovery before the legal principal is written down.
For an insurance sponsor applying IFRS, reinsurance contracts held are analyzed according to IFRS 17's reinsurance requirements, where the applicable contract qualifies. The sponsor's reporting position and the noteholder's financial-asset accounting cannot simply be mirrored because they arise under different legal obligations and potentially different recognition criteria. The related cover may in some structures instead be a derivative or another arrangement whose classification must be separately determined. The present contract analysis identifies economically relevant cash flows but does not purport to adjudicate each entity's financial-statement measurements.
A reconciled closing statement
A comprehensive closing file should connect four ledgers: event and evidence; contractual amount; principal and coupon; and actual cash movement. The event ledger carries peril, occurrence date, observation period, geographic qualification, trigger type, data version, event report, and any challenge. The contractual ledger carries cumulative payouts, prior accepted amounts, deductible/attachment state, limit state, and current payout determination. The securities ledger carries class nominal, each effective reduction, partial repayments, record-date holdings, residual claims, and coupon base. The cash ledger carries collateral realization, eligible sponsor payments, investor principal distributions, costs where contractually permitted, banking dates, and confirmations.
The closing identity for a simplified single-class bond is: initial nominal = total principal reductions + total principal repayments + remaining nominal. A final no-balance security has remaining nominal zero, but it does not follow that all initial nominal was paid to the sponsor: some may have been repaid to investors, some allocated to other contractual settlements, or reduced by a mechanism requiring independent explanation. The beneficiary's receipts are a separate reconciliation. For a conventional fully collateralized, single-risk note without costs or exceptions, catastrophe-linked principal reduction and the sponsor's funded claim may correspond economically; the equality cannot be presumed across all note structures or reporting dates.
Integrated Comparative Synthesis and Research Implications
The observable settlement variable determines the proof burden
The contract chooses what must be established. Indemnity asks whether the sponsor incurred an eligible defined loss, normally requiring claims and adjustments. Industry loss asks whether a particular third-party aggregate passed a threshold at a specified observation vintage. Modeled loss asks what a specified model produces from defined event parameters and exposures. Parametric settlement asks what qualifying physical or geographic observations imply under the payout formula. This classification explains the divergent evidence and time requirements without ranking any trigger as universally superior. The OECD institutional comparison of triggers establishes the underlying variation.
For the sponsor, the relevant performance variable is often how well actual retained losses or emergency funding needs align with the payout; for investors, it is the distribution of principal loss and time to final release; for a regulator, enforceability, funding, and recognition; for an event agent, completeness and reproducibility of the inputs. A measurement convention advantageous to one participant can create a cost for another. For example, closing a parametric result quickly improves liquidity but may embed basis risk relative to claims development. Extending the note until losses mature preserves sponsor protection while restricting the investor's capital availability. These are contract-design consequences, not newly invented institutional doctrines.
Three observed outcomes and their distinct inferential limits
The Mexico 2017 case establishes a full payout on one specified earthquake class; Peru 2019 establishes a 30% payment relative to a $200 million referenced bond; Jamaica 2025 establishes an independently determined 100% payout on its $150 million 2024 parametric note and a documented beneficiary receipt date. Taken together, the examples demonstrate that principal-linked losses can be zero, partial, or complete depending on the event measurement and bond terms, and that post-event cash receipt follows the determination process. These cases are selected because official records are available. They do not constitute a representative statistical sample of all cat bonds or a causal test of different trigger technologies.
The observed Jamaica payout is especially instructive because its event determination notice, its post-event payout case study, and the May 2026 replacement issuance form a traceable sequence. The event was on October 28, 2025; the payout determination was announced November 7; beneficiary cash receipt was on December 1; the later note with $200 million coverage was issued in May 2026. The case shows a realized contingent loss and subsequent fresh capitalization, not a perpetual impairment carried from the earlier security into the new one.
What an actuarial settlement model should output
A useful stochastic settlement analysis extends catastrophe frequency–severity modeling into time and legal state. Its raw simulation draws perils, geographies, and measured losses or hazard parameters. The contract transforms the observations into event eligibility and contingent payoffs. A reporting model simulates notice, measurement vintages, possible revisions, and verification lags. A legal workflow applies cutoffs, objection windows, and extension elections. A securities ledger then computes date-specific nominal reductions, repayments, coupons, and residual claims. This is an analytical decomposition of familiar methods; specifying an integrated workflow does not establish a proprietary actuarial model or a newly discovered statistical estimator.
For a single note, the expected amount of principal loss is E[ΣDk], while expected capital-at-risk duration requires time-indexed outstanding or restricted nominal. A simple exposure-duration summary is E[∫H(t)dt], where H(t) is held capital over a defined horizon. These functionals need different inputs. Two instruments can have the same expected principal reduction yet distinct capital-lockup outcomes. A one-period expected-loss estimate will not capture the distribution of release dates. Any Monte Carlo implementation should specify loss dependence, reporting-lag distributions, information revisions, insolvency/collateral shocks, and correlations among these components where supported; missing data should be represented through explicit sensitivity ranges rather than calibrated to fabricated market frequencies.
Risk measures should be labeled by what they count. Annual expected loss is a risk statistic over a specified exposure horizon, not a guaranteed annual percentage reduction. Attachment probability is the probability of crossing the defined attachment condition, and exhaustion probability is the probability of reaching full payout under the stated assumptions. The World Bank 2024 Jamaica risk summary reports expected loss of 1.61% per annum, attachment probability of 2.46% per annum, and exhaustion probability of 0.84% per annum for that transaction. These issue-level figures should not be interpreted as retrospectively predicted Hurricane Melissa's exact date, or used as market-wide probability estimates.
Settlement latency as a component of effective protection
A sponsor values how quickly it can access catastrophe funding as well as the nominal protection amount. A $150 million payout received one month after a catastrophe and an identically sized payout received after years of underlying claim development create different liquidity profiles. The former may be more effective in financing emergency procurement; the latter may match indemnity more closely. The World Bank's Jamaica recovery account places the $150 million catastrophe-bond payment within a broader set of layered disaster-financing instruments activated after Melissa, with different cash-access mechanics.
An investor similarly values not only the expected amount of loss but the timing of release and possible inability to sell at a stable price during the uncertainty interval. A study that evaluates catastrophe-bond performance solely by eventual principal recovery omits these liquidity effects. Comparing different instruments requires harmonized dates: the qualifying event, determination of amount, ability to transfer economic interest, and actual payment. Without such harmonization a fast public announcement may appear equivalent to a completed cash settlement even when it precedes the latter.
Market pricing, contract specificity, and empirical limits
The risk spread established at primary issuance summarizes investor compensation for a portfolio of perceived risks and current market conditions. Cummins's market development assessment situates catastrophe securities within risk-bearing capacity and reinsurance competition, and Braun's pricing evidence demonstrates that expected loss is influential but not exhaustive. Settlement costs and lockup uncertainty are plausible economic considerations, yet neither publication proves a universal, separately priced 'extension premium' for all securities. Such a coefficient would require transaction-level issue pricing, reported contract extensions, instrument-level controls, and a defensible identification strategy.
The public case record also limits inference about investor distribution. The World Bank reports fifteen global investors in the 2024 Jamaica issue, and Jamaica's 2026 Ministry announcement reports twenty-five in the 2026 issue. Neither statement supplies verified individual allocations at the 2025 record date, secondary-market transfers, or holder-specific realized returns. Pro rata allocation can be derived algebraically from the note conditions without pretending to identify those holders or their ultimate portfolio profits and losses.
Regulatory and public-policy interpretation
Sovereign catastrophe bonds are embedded in broader financing strategies. A full bond payout is a transfer from risk-bearing investors, not a measure of all national disaster damage. It is therefore entirely possible for a country to incur disaster losses greatly exceeding the bond coverage and simultaneously obtain precisely the contractually promised protection. The IMF's 2026 Jamaica program analysis distinguishes the catastrophe-bond fiscal inflow from CCRIF insurance receipts, reserve resources, and debt-creating contingent credit facilities. This distinction matters for evaluating protection effectiveness and public debt consequences.
The investment side presents a different policy question: how can investors and supervisors confirm that risks are funded and disclosures are faithful to the contractual payout? The Solvency II fully funded requirements specify a European supervisory benchmark for qualifying SPVs. The IBRD's notes reflect another institutional route. A common economic description of catastrophe securitization can coexist with heterogeneous legal regimes; a research claim about an investor's enforceable rights must therefore be tied to its actual issuer, note, and legal system.
Research priorities arising from the evidence
Three empirical priorities follow from this synthesis. First, transaction research should construct event–report–payment timelines from authenticated records rather than treating an event press release as cash settlement. Second, comparative performance research should match the definition of loss and the observation period across indemnity, industry-index, modeled-loss, and parametric bonds, explicitly controlling for basis risk and reporting lag. Third, studies of investor returns should measure contingent collateral restrictions and actual noteholder cash flows alongside principal write-downs and quoted prices. Each priority is an evidence requirement for an existing research domain. None claims discovery of a new settlement theory.
Conclusion
Catastrophe bond settlement is a bounded sequence of proof, contractual calculation, allocation, payment, and final discharge. The insured or physical catastrophe gives rise to a potential obligation; the terms of the note and related risk-transfer contracts determine the legally eligible measurement, the resulting payout, and the effect on principal. Principal loss is class-specific and date-specific. Cumulative event reports must be translated into increments without duplication, and effective reductions are capped by outstanding nominal. Pro rata allocation applies within a class when prescribed, whereas different classes and layers may experience different outcomes from the same event.
The 2024 Jamaica CAR 136 documentation establishes that apparently technical clauses—the risk-period definition, calculation-agent event report, five-business-day reporting cutoff, three-business-day manifest-error notice, scheduled principal-reduction dates, conditional maturity extensions, and reduced extension-period risk margin—are essential components of economic risk transfer. The country's October 2025 experience demonstrates full contractual impairment followed by December 1 cash receipt and early redemption. Peru's 2019 partial payout and Mexico's 2017 Class A full payout provide complementary observed outcomes under their own defined structures. These examples establish contractual diversity rather than universal settlement times or payout formulas.
The analysis leads to a practical scientific conclusion: the measurement of catastrophe-bond performance must keep event severity, sponsor loss, trigger payout, principal impairment, market valuation, reserve restriction, and realized cash flow in separate but reconciled records. This separation enables precise actuarial comparison, legally coherent investor allocation, reliable accounting interpretation, and defensible evaluation of sovereign disaster protection. The completed study synthesizes verified public contract terms, official outcomes, supervisory provisions, and scholarly research, with mathematical examples explicitly labeled hypothetical. Its limitations are the absence of confidential transaction cash ledgers, proprietary event model files, full beneficial-owner histories, and independent external peer review; it makes no claim to have tested those inaccessible materials.
References
Braun, A. (2016). Pricing in the primary market for cat bonds: New empirical evidence. The Journal of Risk and Insurance, 83(4), 811–847. https://doi.org/10.1111/jori.12067.
Cummins, J. D. (2008). CAT bonds and other risk-linked securities: State of the market and recent developments. Risk Management and Insurance Review, 11(1), 23–47. https://doi.org/10.1111/j.1540-6296.2008.00127.x.
Cummins, J. D., & Weiss, M. A. (2009). Convergence of insurance and financial markets: Hybrid and securitized risk-transfer solutions. The Journal of Risk and Insurance, 76(3), 493–545. https://doi.org/10.1111/j.1539-6975.2009.01311.x.
European Insurance and Occupational Pensions Authority. (n.d.). Solvency II Single Rulebook: Definitions of special purpose vehicles, Article 13. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2097_en.
European Insurance and Occupational Pensions Authority. (n.d.). Solvency II Single Rulebook: Special purpose vehicles, Directive 2009/138/EC, Article 211. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2298_en.
European Insurance and Occupational Pensions Authority. (n.d.). Solvency II Single Rulebook: Solvency requirements, Delegated Regulation (EU) 2015/35, Article 326. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2615_en.
Federal Reserve Bank of Chicago. (2018). Catastrophe bonds: A primer and retrospective. Chicago Fed Letter, No. 405. https://www.chicagofed.org/publications/chicago-fed-letter/2018/405.
IFRS Foundation. (n.d.). IFRS 9 Financial Instruments. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-9-financial-instruments/.
IFRS Foundation. (n.d.). IFRS 17 Insurance Contracts. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-17-insurance-contracts/.
International Bank for Reconstruction and Development. (2024, April 25). Prospectus supplement: US$150,000,000 floating-rate catastrophe-linked capital at risk notes due December 29, 2027 (Jamaica CAR 136). https://thedocs.worldbank.org/en/doc/2744f96ba4d88e82921b654480991fe4-0340022024/original/ft2024-XS2806459314.pdf.
International Monetary Fund. (2026). Jamaica: Request for purchase under the Rapid Financing Instrument—Press release, staff report, and statement by the Executive Director for Jamaica. IMF Country Report, 2026/008. https://www.elibrary.imf.org/view/journals/002/2026/008/article-A001-en.xml.
Ministry of Finance and the Public Service, Government of Jamaica. (2024, May 3). World Bank settles catastrophe bond issued on behalf of the Government of Jamaica. https://www.mof.gov.jm/world-bank-settles-catastrophe-bond-issued-on-behalf-of-the-government-of-jamaica/.
Ministry of Finance and the Public Service, Government of Jamaica. (2026, May 27). Jamaica secures US$200 million in hurricane insurance coverage through new World Bank catastrophe bond. https://dmb.mof.gov.jm/jamaica-secures-us200-million-in-hurricane-insurance-coverage-through-new-world-bank-catastrophe-bond/.
National Association of Insurance Commissioners. (2025). Insurance-linked securities. https://content.naic.org/insurance-topics/insurance-linked-securities.
Organisation for Economic Co-operation and Development. (2024). Fostering catastrophe bond markets in Asia and the Pacific. OECD Publishing. https://www.oecd.org/en/publications/fostering-catastrophe-bond-markets-in-asia-and-the-pacific_ab1e49ef-en.html.
Organisation for Economic Co-operation and Development. (2024). Fostering catastrophe bond markets in Asia and the Pacific: Overview of structures and triggers [Online chapter]. https://www.oecd.org/en/publications/fostering-catastrophe-bond-markets-in-asia-and-the-pacific_ab1e49ef-en/full-report/component-5.html.
Organisation for Economic Co-operation and Development. (2024). Fostering catastrophe bond markets in Asia and the Pacific: Recent trends and characteristics [Online chapter]. https://www.oecd.org/en/publications/fostering-catastrophe-bond-markets-in-asia-and-the-pacific_ab1e49ef-en/full-report/component-6.html.
Organisation for Economic Co-operation and Development. (2024). Fostering catastrophe bond markets in Asia and the Pacific: Benefits and challenges [Online chapter]. https://www.oecd.org/en/publications/fostering-catastrophe-bond-markets-in-asia-and-the-pacific_ab1e49ef-en/full-report/component-7.html.
Organisation for Economic Co-operation and Development. (2024). Fostering catastrophe bond markets in Asia and the Pacific: Country case studies [Online chapter]. https://www.oecd.org/en/publications/fostering-catastrophe-bond-markets-in-asia-and-the-pacific_ab1e49ef-en/full-report/component-8.html.
World Bank. (2021, November 5). Advancing policy reforms in Peru to reduce risk from natural hazards. https://www.worldbank.org/en/results/2021/11/05/advancing-policy-reforms-in-peru-to-reduce-risk-from-natural-hazards.
World Bank. (2024, April 25). World Bank returns to the cat bond market providing financial protection to Jamaica. https://www.worldbank.org/en/news/press-release/2024/04/25/world-bank-returns-to-the-cat-bond-market-providing-financial-protection-to-jamaica.
World Bank. (2025, November 7). Hurricane Melissa triggers 100% payout of $150 million World Bank catastrophe bond for Jamaica. https://www.worldbank.org/en/news/press-release/2025/11/07/hurricane-melissa-triggers-100-payout-of-150-million-world-bank-catastrophe-bond-for-jamaica.
World Bank. (2025, December). World Bank catastrophe bond renews $150 million hurricane coverage for Jamaica: Case study (updated following Hurricane Melissa). https://thedocs.worldbank.org/en/doc/401877c87631461af8ad227793affc5f-0340012025/original/Case-Study-Jamaica-2024-Cat-Bond.pdf.
World Bank. (2025, December 17). How Jamaica's preparedness delivers after a devastating hurricane. https://www.worldbank.org/en/news/feature/2025/12/17/jamaica-preparedness-delivers-after-devastating-hurricane.
World Bank. (2026, May 18). World Bank prices catastrophe bond providing protection for Jamaica replacing coverage triggered by Hurricane Melissa. https://www.worldbank.org/en/news/press-release/2026/05/18/world-bank-prices-catastrophe-bond-for-disaster-protection-for-jamaica.
World Bank. (2026, June). Third catastrophe bond for Jamaica: Increased cover following Hurricane Melissa. https://thedocs.worldbank.org/en/doc/f66f98e820b44dcec2070273327df145-0340012026/original/Case-Study-Jamaica-2026-Cat-Bond.pdf.
World Bank Treasury. (n.d.). Capital at risk notes. https://treasury.worldbank.org/en/about/unit/treasury/ibrd/ibrd-capital-at-risk-notes.