ENSO does not cause insurance losses. It tilts the odds of the weather that causes them — and because insurance is priced and reserved on expected frequency, a tilt in the odds is enough to move underwriting results, reinsurance pricing, and the catastrophe loads that sit inside an insurer’s combined ratio. This note traces that chain from the equatorial Pacific to the income statement, for the two perils where the US signal is clearest: Atlantic hurricanes and severe convective storms.

01 Two perils, opposite polarities

The single most important thing to get right — and the thing most casual commentary gets wrong — is that ENSO pushes the two major US catastrophe perils in opposite directions. There is no single “El Niño is bad for insurers” statement that holds. It depends entirely on which peril and which coast.

One critical caveat the research itself stresses: the ENSO–SCS signal is strongest in spring and weakest in summer — and summer is when severe storms actually peak across much of the country. So the predictive signal is real but seasonally bounded. An analyst leaning on ENSO for SCS frequency is borrowing a March–May edge, not a year-round one.

02 Atlantic, Gulf, and Pacific — geography of the signal

The same ENSO state expresses differently by region, because the physical mechanism differs by region.

Atlantic & Gulf of Mexico — the wind-shear story

El Niño increases vertical wind shear across the tropical Atlantic and Caribbean. Shear is hostile to hurricane formation: it disrupts the vertical column a storm needs to organise. So El Niño years tend to produce fewer and weaker Atlantic and Gulf hurricanes, La Niña years more and stronger. For insurers this is the highest-severity channel — a single major Gulf or Florida landfall can dwarf an entire season of convective losses — which is why the hurricane signal gets the most attention even though, on an average annual basis, convective storms now rival it.

Southern Plains & Midwest — the convective story

This is the SCS belt — Tornado Alley and the hail corridor running from Texas up through Kansas, Nebraska, and into the Midwest. Here the La Niña-enhances / El Niño-suppresses spring relationship governs. The losses here are higher-frequency, lower-individual-severity than hurricanes, but they aggregate: hail alone is the quiet majority of SCS insured losses, even though tornadoes get the headlines.

Pacific coast — the indirect story

The US Pacific coast has little direct hurricane or tornado exposure, so ENSO’s insurance relevance there runs through precipitation and wildfire-adjacent pathways: El Niño tends to wet California (flood, mudslide, and debris-flow exposure after burn scars), while La Niña tends to leave the Southwest drier (compounding multi-year drought and the wildfire fuel load). The 2025 California wildfire losses (~$40bn) are a reminder that the Pacific channel, while not a “storm” peril, is now a first-order driver of US catastrophe totals.

03 From Pacific SST to the income statement

The mechanism that matters to an analyst is the chain from a sea-surface-temperature anomaly to a number in a financial statement. It runs like this:

The key conceptual point: insurers do not price the actual storms of a given year — those are unknowable at the point of pricing. They price the expected loss distribution, expressed through catastrophe models as an Average Annual Loss (AAL) and a set of return-period losses (the 1-in-100, 1-in-250). ENSO enters by shifting the near-term event-rate assumptions inside those models away from the long-run climatological average.

04 Where it lands: P&C primary vs reinsurance

The two sides of the market absorb the ENSO signal very differently, because they sit at different points in the loss tower.

05 How this is tracked — and why it matters

ENSO is one of the few catastrophe drivers that is observable months ahead. Hurricanes and tornadoes are not forecastable individually at a seasonal horizon, but the ENSO state that tilts their odds is monitored continuously by NOAA, IRI, ECMWF, and the BOM. That lead time is exactly what makes it useful for an industry that has to set prices and buy reinsurance before the season it is exposed to.

What analysts actually watch

The practical toolkit: the ENSO forecast plume and official advisories (for the phase heading into wind season); the major cat-model vendors’ seasonal and “near-present” event-rate views (Verisk and Moody’s RMS); the broker seasonal outlooks (Gallagher Re, Aon, Guy Carpenter); and the running tally of catastrophe losses through the year from the same brokers, which tells you how much of the aggregate covers has already eroded. For equity analysts covering carriers, the read-through is to the cat load in guidance, the pace of reserve development, and management commentary on reinsurance cost at renewal.

Why tracking weather events becomes a financial discipline

For an insurer, the ENSO signal is not used to predict a specific storm — it is used to set the central assumption around which capital is allocated. A La Niña-tilted outlook can justify a higher cat load, an earlier or larger reinsurance purchase, a more conservative IBNR posture, and more cautious earnings guidance. If the season then under-delivers, that conservatism shows up as favourable reserve development (a release); if it over-delivers, the pre-positioning is what prevents a capital event. The discipline is symmetric: the point of tracking is to be neither over- nor under-reserved relative to the tilted odds.

06 The bottom line

ENSO is a probability tilt, not a forecast. Its value to insurers and their analysts is that it is observable ahead of the exposure period, it points the two big US perils in opposite directions, and it lands differently on primaries (frequency, cat load, IBNR) than on reinsurers (severity, aggregate erosion, renewal pricing). The 2026 setup — a developing El Niño — argues for hurricane suppression into the back half of the season and a quieter spring convective signal, but with the standing caveats that one landfall rewrites the hurricane math and that the convective signal fades into the summer peak. For anyone modelling a carrier, the ENSO state is best treated as an input to the cat-load and reinsurance-cost assumptions, not as a directional call on any single quarter’s results.