Insurance Desk

Cat Bond Desk Soren Vaeth

Let's read the tape. YTD issuance of $18.9B across 92 deals, outstanding risk capital at $65.6B, market yield 9.29% — of which 5.53% is pure insurance risk spread against a market-level expected loss of 2.5%. That puts the multiple-on-EL at roughly 2.2x. That is not a distressed market begging for capital; that is a functioning, price-discovered market where investors are being compensated materially above the actuarial burn rate. The collateral yield component — 3.76% on T-bill or money-market collateral at effective fed funds of 3.63% — is essentially free money layered on top of the risk premium. This is why the pipeline has not dried up.

The July deal cluster is instructive about where the stress is concentrated. The 123 Lights Re transaction — $100M, cedent the Los Angeles Department of Water and Power, peril California wildfire — is a municipal utility buying capital-markets protection for wildfire ignition liability. That is a different risk profile than a homeowner policy: it is a concentrated, correlated, legally-exposed liability sitting on a public balance sheet. The fact that LADWP can place $100M in the ILS market tells you two things: investor appetite for California wildfire risk exists at a price, and the price is being set here rather than in the admitted market, which has largely exited the state. The $345M Matterhorn Re (Swiss Re) and $200M 3264 Re (Hannover Re) named-storm and earthquake structures show the major Bermuda and European reinsurers continuing to offload peak-peril exposure to the capital markets at current spreads.

The average recent deal size of $138M and the Porch Group's $100M Harbor Crest Re structure — covering named storm, winter storm, severe weather, wildfire, and fire-following earthquake for a technology-enabled homeowners carrier — are both worth flagging. Porch accessing the cat-bond market for multi-peril protection signals that even mid-tier primary carriers are finding ILS a viable complement to or substitute for traditional reinsurance towers. As long as the HY OAS sits at 2.7% (tight, risk-on) and the VIX holds at 15.19, relative-value money will keep flowing into this asset class. The risk I track but sometimes underweight: if a major Atlantic named storm makes landfall before October 1, trapped capital in named-storm tranches could spike, and that LADWP wildfire bond's recovery assumptions would be tested by a simultaneous California fire event in the same loss year.

At a 2.2x multiple-on-EL and 9.29% market yield, ILS investors are well-compensated; the LADWP wildfire bond and Porch Group multi-peril deal signal that capital-markets risk transfer is filling the void left by admitted-market retreat from California and complex multi-peril books.

Bias flag — Reads the 2.2x multiple-on-EL as adequate compensation; structurally underweights the scenario where model error on California wildfire ignition liability or SCS aggregate loss is large enough that the EL itself is understated, making the multiple illusory.

The Cycle Margaret Ennis

Soren's arithmetic on the multiple-on-EL is clean, but I'd push back on the conclusions he draws from it. A 2.2x multiple-on-EL sounds like adequate compensation — and in a stationary world it is. The cycle lens says something different: $18.9B of YTD issuance, 92 deals, an average size of $138M, and a pipeline that includes Swiss Re, Hannover Re, Porch Group, and LADWP all printing in July alone tells me capital is not scarce. Capital is abundant. And abundant capital, in the reinsurance cycle, is the precondition for the next softening. We have been here before. The post-Katrina hardening attracted Bermuda Class of 2005/2006 capital. The post-Ian/Ida hardening attracted the current ILS expansion. At some point — and the $65.6B outstanding market is the tell — supply catches up with demand and rate-on-line begins to slide.

The Descartes parametric SCS piece is the day's most structurally important piece of news for the cycle, even though it received the least market attention. SCS is the peril that has been quietly driving combined ratios above 100 for regional carriers and generating loss creep that traditional aggregate reinsurance structures were not designed to absorb. Descartes' argument — that parametric, index-based solutions are more responsive for aggregate SCS exposure than indemnity structures — is a cycle argument as much as a product argument. If parametric SCS structures gain traction, they shift risk transfer from opaque, lag-ridden indemnity settlement into transparent, rapid-trigger instruments. That changes the information environment for reinsurers pricing the aggregate layer. Whether that ultimately tightens or loosens the cycle depends on whether parametric triggers are set with enough basis risk that cedents remain exposed, or tight enough that they actually transfer the loss. The market is not there yet. Watch whether the ILS market begins pricing SCS parametric deals the way it now prices named-storm parametric deals — that transition, if it happens, will be visible in the Artemis deal directory before it shows up in renewal ROL data.

Record-pace ILS issuance signals capital abundance, not scarcity — the precondition for the next softening; parametric SCS structures, if they scale, could reshape how aggregate convective-storm risk is priced and transferred across the cycle.

Bias flag — Mean-reversion lens may miss that this ILS cycle is structurally different: the collateral yield tailwind (3.76% on T-bills at 3.63% fed funds) is suppressing the risk premium floor in a way that did not exist in prior soft markets, making 'capital abundance leads to softening' a less automatic prediction.

Modeled Loss Dr. Ravi Chandrasekar

The Descartes SCS research is the empirical anchor for today's session. The finding — no sign of reversal in SCS intensity or frequency — is consistent with what the peril modeling community has been documenting for the better part of a decade: severe convective storm loss is not mean-reverting in the way that Atlantic hurricane loss has historically been modeled. SCS events are geographically dispersed, high-frequency, and have a correlation structure that is poorly captured by single-event EP curves. The aggregate exposure problem — many moderate events accumulating into a catastrophic annual loss — is exactly what traditional occurrence-based reinsurance structures were not designed to handle. This is not a model failure in the catastrophe-model sense; it is a structural mismatch between the loss distribution and the contract form.

The LADWP wildfire cat bond is the other modeling story today. California wildfire ignition liability for a utility is a genuinely hard modeling problem. The loss is not just a function of fire weather and fuel load — it is a function of ignition probability from infrastructure, legal doctrine (inverse condemnation in California means a utility can be liable even without negligence), and demand surge in post-fire reconstruction. The cat-bond market is pricing this, but I would want to see the attachment probability and exhaustion probability disclosed before drawing comfort from the $100M placement. A bond that attaches at the 0.5% annual exceedance probability is very different from one that attaches at 5%. The corpus does not give us those numbers, so I flag the gap. What the LADWP deal does confirm is that the modeled loss is credible enough for institutional investors to take the other side — that is a meaningful signal, even without the granular EP curve.

The New Zealand climate-tort bill is worth a brief actuarial note. Closing domestic courts to climate-harm claims does not eliminate the underlying physical loss — it eliminates one legal pathway to recovery. For reinsurers and ILS investors, this is mildly positive: it reduces the litigation-driven loss amplification that no peril model captures. But it is a single jurisdiction. The structural trend — courts in common-law jurisdictions increasingly entertaining attribution science as a basis for tort liability — remains intact globally. New Zealand is the exception, not the trend.

SCS loss non-stationarity makes aggregate exposure the hardest modeling problem in property cat; the LADWP wildfire bond is a meaningful market signal, but the corpus lacks the attachment and exhaustion probability data needed to assess whether it is priced correctly.

Bias flag — Treats institutional investor willingness to buy the LADWP bond as model-credibility evidence; underweights the possibility that investors are yield-chasing in a tight HY spread environment (OAS 2.7%) and that the model is less tested than the placement implies.

Solvency Watch Eleanor Pryce

The LADWP wildfire cat bond is, from a regulatory standpoint, the most interesting transaction in today's corpus — and not for the reasons Soren finds interesting. LADWP is a municipal utility, not a regulated insurer. It does not file with a state department of insurance. It does not hold RBC capital. Its wildfire ignition liability sits on a public balance sheet backstopped implicitly by Los Angeles ratepayers. The $100M ILS placement is, in effect, LADWP using the capital markets to offload a regulatory-capital-free liability that a regulated insurer would have to reserve against explicitly. There is no AM Best rating action that covers this. There is no NAIC RBC charge that captures it. This is a solvency-adjacent story that falls through the regulatory cracks — and as California wildfire losses continue to accumulate, the question of whether LADWP's unregulated liability exposure gets properly ring-fenced matters enormously for the CA FAIR Plan and the broader California admitted market.

The New Zealand climate-tort bill is a direct regulatory-solvency story. By closing courts to climate-harm claims, New Zealand's parliament is making a legislative determination that certain climate-attributable losses will not be recoverable through litigation. For New Zealand-domiciled insurers and reinsurers, this reduces tail liability. For global reinsurers with New Zealand exposure, it is a modest positive. But ClientEarth's warning about precedent is the more consequential signal: if other common-law jurisdictions follow, the litigation backstop that has historically amplified insured losses (and driven reserve development) gets narrowed. That would be structurally deflationary for loss reserves — which sounds good until you realize that loss reserves are also how regulators know whether a carrier is solvent. A regime with less litigation-driven reserve development is also a regime where solvency problems may surface more slowly and later.

LADWP's wildfire cat bond exposes a regulatory blind spot — municipal utility wildfire liability sits outside RBC and NAIC oversight, yet its failure would cascade into the CA FAIR Plan; New Zealand's climate-tort closure is a reserve-development signal as much as a liability one.

Bias flag — Reads LADWP's off-balance-sheet liability structure as primarily a regulatory blind spot; underweights the possibility that the cat-bond placement is a net positive for California ratepayer solvency risk by ring-fencing some ignition liability in capital-market structures.

Protection Gap Daniela Owusu-Reyes

The LADWP wildfire cat bond is a $100M solution to a multi-billion-dollar problem — and I want to be precise about what problem it does and does not solve. It protects LADWP's balance sheet against wildfire ignition liability. It does not protect a homeowner in Pacific Palisades or Altadena who cannot get admitted-market coverage, cannot afford CA FAIR Plan premiums, and is simultaneously watching their insurer-of-last-resort tighten its coverage terms. The capital markets are solving for institutional balance-sheet risk. The protection gap — the gap between economic loss and insured loss — keeps widening for the households at the end of that supply chain.

The Descartes SCS parametric piece is where I find the most consumer-relevant signal today, and it cuts both ways. Parametric insurance, properly designed with narrow basis risk, can pay faster and more transparently than indemnity insurance after a severe convective storm. For low-income households and small businesses in tornado alley and the Gulf Coast interior — populations that have historically been underinsured against SCS because the peril was considered too granular for cat modeling — parametric triggers could improve coverage access. The problem is distribution: parametric products currently reach sophisticated commercial buyers and agricultural policyholders in development-finance contexts (the Nepal crop insurance story in today's corpus, though in a very different market, illustrates this gap). The retail household market in Tulsa or Memphis does not have access to a Descartes parametric SCS product. Until that distribution problem is solved, 'parametric can help on aggregate SCS exposure' is a reinsurer and cedent story, not a consumer story. The protection gap for SCS events — frequent, dispersed, and systematically underinsured in personal lines — is the one Dr. Chandrasekar's EP curve is least equipped to measure, and the one that shows up most painfully in the economic loss vs. insured loss divergence after every major tornado outbreak.

The LADWP wildfire cat bond protects an institutional balance sheet, not households; parametric SCS solutions exist for sophisticated buyers but the retail distribution gap means the personal-lines protection gap for severe convective storms remains wide and growing.

Bias flag — Frames the entire LADWP transaction and parametric SCS discussion as institutional solutions that fail consumers; underweights the pathway by which institutional ILS structures lower reinsurance costs that eventually flow through to primary-market pricing for households.