Cat Bond Spread Compression Tests Retro Pricing

Cat bond risk spreads averaged 5.72% in late May 2026, 13% below the 6.59% level of a year earlier (Plenum, May 2026). Everest Group's Kilimanjaro III Re retrocession transaction upsized to $630 million and priced all six tranches at the bottoms of already-reduced guidance (Artemis, June 2026). When a deal grows and still clears below the floor of marketing ranges, the market is setting the marginal cost of catastrophe capital, not negotiating around it.

That distinction carries weight for cedents designing retrocession programs. The transaction-level data from Kilimanjaro III Re is not a pricing anecdote; it is one entry in a sequence of trades that together constitute a real-time yield curve for catastrophe risk transfer. Reading that curve accurately determines whether the current environment justifies substituting cat bonds for indemnity retro, whether the multi-year tenor of bonds is a strategic lock-in opportunity or a liquidity trade-off, and whether the investor appetite now visible across North America storm, European flood, and other peril classes is durable or one active season away from reversal.

The Kilimanjaro Transaction: Upsize and Below-Guidance Clearing

Everest initially brought Kilimanjaro III Re to market targeting $530 million across two series. Investor demand pushed the ceiling to $675 million before the deal settled at $630 million. Both series cleared at the bottoms of guidance ranges that had already been reduced during bookbuilding: Series 2026-1 at $350 million over three years and Series 2026-2 at $280 million over four years, together covering North America named storms and earthquakes affecting the U.S., Puerto Rico, U.S. Virgin Islands, Washington D.C., and Canada on a per-occurrence and annual aggregate basis across six tranches, with final pricing spanning 6.75% to 11.75% (Artemis, June 2026).

Post-settlement, Everest holds $1.83 billion in cat bond protection outstanding, with no maturities until 2028 (Artemis, June 2026). That structure has actuarial relevance independent of the pricing. A multi-year capital stack with no near-term renewal exposure insulates Everest from a hard-market turn: if La Nina conditions return in 2027 and Atlantic activity spikes, Everest's retrocession cost through 2029 is already fixed. The three-year and four-year tenors on Kilimanjaro III Re are not incidental features; they are the strategic payoff of transacting in a soft cat bond market.

The pattern across the deal's lifecycle is the signal. Guidance reduced; size grew; pricing cleared at guidance floors. That three-part sequence tells you investor demand was sufficient to absorb a $100 million upsize while still compressing to the low end of a range that had already moved in buyers' favor. A transaction that closes at target on unchanged guidance is a neutral data point. Kilimanjaro III Re is directional evidence that the clearing price sits below the marketed range, which puts the actual marginal cost of capital to Everest lower than the 6.75% to 11.75% band suggests.

Retrocession Instruments: Basis Risk, Collateral, and Renewal Flexibility

Cedents comparing instrument types in this environment are navigating three variables simultaneously: basis risk, collateral structure, and renewal flexibility. Each instrument optimizes differently across that space, and the current ILS pricing environment shifts the relative trade-offs in ways that are not obvious from spread compression alone.

Indemnity retrocession carries no basis risk by construction: the cedent's actual net losses trigger recovery, the same logic as a traditional per-risk or per-occurrence treaty. The trade-off is credit and renewal exposure. Traditional retro is typically unfunded, backed by letters of credit and rated counterparty strength rather than posted collateral, meaning a protracted event development period can slow recovery. More importantly, indemnity retro renews annually, and annual pricing resets to market. A cedent buying indemnity retro at 2026 levels has no protection against a hard-market repricing after a 2027 active loss season.

Industry loss warranties settle against an industry loss index, typically PCS estimates in the U.S. context, rather than the cedent's actual losses. Basis risk is real and, for specialty books with geographic or line concentrations that diverge from industry composition, can be material. The compensation is speed: ILW settlements are faster than indemnity development, pricing is highly transparent because the trigger is public, and the market is liquid enough for secondary trading during active seasons. Like indemnity retro, ILWs are annual instruments and reprice to market at renewal.

Cat bonds are fully collateralized instruments, eliminating counterparty credit risk entirely. The collateral sits in a special-purpose vehicle invested in money market or Treasury instruments; the cedent does not rely on a reinsurer's balance sheet for recovery. Trigger types range from indemnity to parametric to industry index, each trading basis risk against settlement speed. The distinguishing feature in the current market is tenor: cat bonds typically run three to five years, meaning a cedent who transacts now locks in 2026 spread levels through 2029 or 2030 regardless of how the reinsurance cycle moves. At 5.72% average risk spread, that multi-year lock-in is worth pricing carefully against the 2027 renewal cost an active season could impose on annual instruments.

Sidecars are quota-share structures where investors participate in a defined slice of the cedent's own underwriting book, rather than covering a layer above a specified loss threshold. The cedent retains the original underwriting risk profile; the sidecar investor receives a return on line proportionate to actual underwriting results plus investment income on posted collateral. Sidecars reset annually in most structures, and investor appetite for them is sensitive to underwriting returns rather than pure cat risk pricing. They are capital-raising vehicles as much as risk transfer mechanisms; the cedent cedes premium and losses rather than paying a spread-based coupon, making them structurally different from both cat bonds and retro treaties in how they interact with the cedent's balance sheet.

Retrocession Instrument Comparison: Current Market Context

Instrument	Basis Risk	Collateral	Tenor / Renewal	2026 Market Signal
Indemnity Retro	None (actual loss)	LOC / rated counterparty	Annual; market resets at renewal	Pricing tracks cat bond compression; no multi-year price lock
Industry Loss Warranty (ILW)	Material (index vs. actual loss)	Often collateralized	Annual; transparent repricing	Liquid; spreads compressed; fast settlement post-event
Catastrophe Bond	Low to medium (trigger-dependent)	Fully collateralized	3-5 years; public market pricing	Multi-year lock at 2026 lows; Kilimanjaro III Re sets the benchmark
Sidecar	Low (quota share of actual book)	Fully collateralized	Annual; investor-appetite sensitive	Return tied to underwriting results, not pure spread; capital-raising vehicle

The current environment tips the calculus toward cat bonds on the multi-year tenor dimension. Cedents who can accept the basis risk of non-indemnity triggers, or who can access the 144A market for indemnity cat bonds, have an opportunity to lock in compressed spreads through the end of the decade. That opportunity expires at the next major active cat season, when investor demand and supply dynamics can reverse quickly. Everest has already acted. The question for other retrocession buyers is how much of the remaining window they intend to use.

Peril Appetite in 2026: Storm, Flood, and the European Market

Kilimanjaro III Re covered North America storm and earthquake, the two perils that constitute the deepest and most liquid segment of the cat bond market. But 2026 issuance has extended materially beyond those traditional perils, and the spread signals across the expanded peril set carry their own content for cedents evaluating whether ILS substitution is viable outside North America peak-zone exposure.

Through late June 2026, total cat bond issuance reached $16.1 billion year-to-date, on pace for one of the largest first-half totals on record (Artemis, June 2026). The mix includes European sponsors bringing peak-peril risks to the 144A market, new geographic territories, and flood-specific transactions that would have been difficult to place in earlier years.

Gothaer Versicherungsbank's Yardstick Re transaction illustrates where appetite has extended. The EUR 100 million debut cat bond, four years in tenor, covered German river flood losses on an indemnity basis and priced at 1.95%, at the low end of the 1.95% to 2% guidance range (Artemis, June 2026). The notes carry an initial expected loss of 0.19%. The spread-to-expected-loss multiple is therefore approximately 10.3 times, meaning investors received roughly ten dollars of risk premium for every dollar of long-run expected loss. That multiple reflects the premium demanded for model uncertainty, illiquidity relative to public equity markets, and event concentration risk; it also tells you that even at 1.95%, European flood ILS is compensating investors generously for the underlying risk relative to its nominal spread level.

The contrast with North America peak-peril pricing is instructive. Kilimanjaro III Re's highest-attachment tranches cleared around 6.75% on underlying expected losses that for remote layers likely run well below that figure; the spread-to-EL multiple for remote catastrophe bond tranches regularly exceeds five to seven times. European flood at Gothaer's 1.95% on 0.19% EL is paying a higher multiple on a lower absolute spread. Both are confirming that investor demand is sufficiently strong that even perils with thin loss history in the ILS market are clearing at multiples that reflect risk aversion and model uncertainty rather than just pure expected loss pricing.

Wildfire remains selective. The January 2025 Los Angeles fires injected fresh model uncertainty into California wildfire cat bond pricing, and 2026 issuance in that peril has shown less compression than storm or flood. Investors pricing wildfire in 2026 are working from post-LA event model recalibrations that have not yet stabilized, and the spread premium for California wildfire reflects that uncertainty. For cedents with meaningful California wildfire exposure seeking ILS-based retrocession, the market is open but the marginal cost stays elevated relative to its pre-event 2024 level.

ILS Spread Compression and the Catastrophe Load Problem

Cat bond secondary market spreads often move as a leading indicator of where traditional treaty pricing will settle. That observation has a corollary: ILS spread compression in 2026 does not represent only a reassessment of underlying risk. It also reflects supply and demand in the capital markets, two years of loss-free experience compounding into strong fund returns, and deployment pressure on institutional ILS allocations. Folding the current 5.72% average spread directly into a primary catastrophe load, as though it represented the long-run expected cost of catastrophe exposure, builds a soft-market assumption into rates that must survive across a full underwriting cycle.

The Plenum data from May 2026 gives the decomposition: the cat bond market's 9.42% total yield consists of a 2.39% average expected loss, a 3.70% risk-free return on collateral, and a 3.33% risk premium above expected loss (Plenum, May 2026). The multiple of risk premium to expected loss is roughly 1.4 times at current market. At the 2023 peak of post-hurricane repricing, comparable transactions ran at multiples of 2.5 to 3 times expected loss, and the risk premium was correspondingly larger. The current multiple signals that the market has given back a significant portion of the post-event hardening, and that the risk premium embedded in current spreads is approaching pre-2022 levels.

For a carrier actuary setting primary catastrophe loads, this distinction matters. The appropriate cat load reflects the long-run expected cost of catastrophe exposure at the vendor model's current parameters, not the market's current supply-and-demand clearing price. If ILS spreads are compressing because institutional capital is overweight the asset class, that is a capital markets condition, not a signal that underlying hazard has diminished. The June 2026 cat renewal analysis documented the actuarial risk in letting reinsurance market signals drive primary cat load selections downward; the ILS spread data reinforces the same concern from the capital markets angle.

For cedents evaluating retrocession structure, ILS spread levels also affect earnings volatility in a way that is sometimes overlooked. Purchasing cat bond retrocession at compressed spreads reduces the cost of protection and, in a benign year, reduces the earnings drag relative to unprotected exposure. The less visible effect is that total recoveries from a retrocession program are fixed at the structure's limit; if primary exposure has grown through premium inflation, expanded territorial appetite, or higher insured values, the same nominal retro limit provides less effective protection against a tail event than it did when the program was sized. Cedents buying retro in a soft market should verify that the program limit reflects the current exposure base, not the base at last major program design, before treating the spread saving as pure margin improvement.

When Investors Pull Back: The Florida Pension Signal

Trent Webster, Senior Investment Officer at the Florida State Board of Administration, told the board's investment advisory council in June 2026 that the Florida Retirement System is considering a reduction in its ILS allocation if the soft market persists. "Cat bonds really aren't that attractive right now," Webster said, noting that the system's three-year ILS return has been 18% but that the forward-looking expected return has been declining as spreads compress (Florida SBA/Artemis, June 2026). The system held approximately $2.23 billion in ILS and reinsurance strategies at year-end 2025, representing about 1% of total pension assets, with the allocation edging to roughly $2 billion and 0.9% of the fund by March 31, 2026.

Webster's comment "We could have a pretty significant reduction if it's a quiet year" identifies the self-limiting mechanism embedded in every soft ILS cycle. Institutional allocators enter the asset class at returns above their required hurdles, scale up aggressively through benign loss periods as realized returns compound, then re-evaluate as prospective returns compress toward those hurdles. The SBA's three-year return of 18% is a backward-looking figure; the forward-looking return implied by a 5.72% average risk spread and a 2.39% expected loss is a 3.33% risk premium above risk-free, which is a materially different proposition from the spread levels at which many pension funds built their initial ILS positions in 2022 and 2023.

The total ILS market outstanding reached approximately $120 billion at year-end 2025 (Gallagher Re, 2026), with pension fund capital representing a large and growing portion of that capacity. If pension allocators broadly reduce target ILS weights in response to compressed expected returns, the market loses the marginal capital that closed the supply-demand gap in 2025 and early 2026. That capital withdrawal can happen quickly: ILS fund redemptions, unlike traditional reinsurance capital, are not locked up for multi-year periods at the institutional level, though individual cat bonds have fixed tenors. The combination of instrument-level tenor lock and investor-level flexibility creates a dynamic where outstanding notional can remain large even as new issuance demand softens; it is the secondary market spread that adjusts first.

The SBA signal matters not only because of its own allocation size but because it articulates a price sensitivity that, if generalized across the pension fund ILS universe, could reverse the supply that compressed spreads in the first place. Most ILS market participants estimate the institutional hurdle rate at somewhere between 6% and 8% total yield at current risk-free rates. The May 2026 market yield of 9.42% is above that range, which explains why the SBA has not yet exited. But the trajectory since late 2025 has been directional toward the hurdle. A quiet 2026 Atlantic season compresses spreads further; further compression reduces prospective yield; reduced prospective yield accelerates institutional exit. The mechanism is self-limiting only in the sense that it eventually reverses: below some floor, the capital that was withdrawn returns at the next major event and the cycle restarts.

Cedents evaluating multi-year cat bond transactions in this environment face a specific timing question. At what point in the spread-compression cycle does the multi-year lock-in benefit exceed the option value of waiting for the next hard market to provide cheaper relative protection against a post-event cost spike? The Florida SBA signal, combined with the Kilimanjaro III Re structure, suggests the answer for most large retro buyers: the optimal window for locking multi-year protection at current spreads is narrowing. Cedents who have not yet transacted are making a bet that the 2026 season is benign enough to compress spreads further before a loss event resets the market. That bet has a payoff, but it is one-sided. The downside of being caught without retrocession in a hard market is asymmetrically larger than the upside of saving a few additional basis points on the coupon by waiting.

Carriers that completed this analysis in the 2015 soft cycle and locked multi-year ILS protection at pre-2017 spreads entered the Harvey-Irma-Maria sequence with their retrocession cost already committed at below-event-level pricing. Those that deferred and lost the window spent the 2018 and 2019 renewal seasons buying protection from a fundamentally different supply position. The Q1 2026 cat bond market analysis documented the institutional inflow cycle that preceded the current spread compression; the Florida SBA comment is the first clear signal that the inflow cycle is approaching its limit.

Further Reading on actuary.info

• June 2026 Cat Renewals Signal 15-20% Pricing Decline and a Growing Model-to-Market Gap: How the June 1 traditional reinsurance renewal confirmed the same softening trajectory that cat bond spreads had telegraphed earlier in the year, with rate-on-line data from Howden Re and Guy Carpenter and the actuarial implications for primary cat load calibration.
• Cat Bond Market Hits $63.9B as Pension Funds Scale Up: Q1 2026: The institutional capital flows that drove spread compression, including UCITS fund inflows, pension allocation growth, and the supply-demand dynamic that set the stage for the Kilimanjaro III Re pricing outcome.
• Cat Bond Issuance Pace 2026: Maturities, Reinvestment, and the Spread Floor: The maturity schedule and reinvestment dynamics in the 2026 cat bond market, including how maturing principal from 2022 and 2023 hard-market issuance is repricing into current spreads.
• Swiss Re Q1 2026: $1.5B Profit Masks a Nat Cat Cycle Pivot: The reinsurer-side view of the same capital cycle, with Swiss Re's retrocession strategy changes, combined ratio trajectory, and what the largest natural catastrophe reinsurer's positioning signals about the durability of current ILS spread levels.
• Zurich's 13-Year Cat Bond Hiatus Ends with Turicum Re: The sponsor-side decision to re-enter the cat bond market after a multi-year absence, including the strategic context and how sponsor re-entry behavior during soft cycles informs retrocession purchasing timing for other cedents.