THE LAYER SWISS RE STILL HASN'T BUILT

Swiss Re Institute published two numbers about the same year. One: natural catastrophe losses in 2025 ran to$220 billion, of which $107 billion was insured, a record 49% share, the highest on the firm's own records. The other: the global natural catastrophe protection gap in 2025 was$424 billion, up from $395 billion the year before. Nearly four times the shortfall the first pair of numbers implies. The same institution, the same year, two different measures of the same problem, and only one of them is the number that gets quoted. The first pair is a subtraction: what happened, minus what was paid. The second is a forecast, built from modelled expected losses and expressed in premium-equivalent terms, measuring what should be covered against what the risk actually requires, not what one calendar year happened to demand. The second number is the real one. It also happens to be the one you have to leave the annual results and go looking through a separate piece of research to find. Break the $424 billion open by region and the shape sharpens. Swiss Re's own Natural Catastrophe Insurance Resilience Index gives Emerging Asia a score of5%. Not five percent of a slower-growing market. Five percent of everything newly built, planted, and financed across the region this year that could be lost to a catastrophe carries cover. The other ninety-five percent carries the loss alone. More broadly, across emerging economies, Swiss Re's own words:80 to 90%of catastrophe losses are still uninsured. Their own prescription for closing it: enhancing theawareness, attractiveness, access and affordability of insurance, four levers, all on the demand side, telling people it exists, making it worth having, putting it within reach, pricing it for a household with no financial reserve to draw on. Nothing in that list is about what the policy actually measures once someone buys it. Anthony Hobley went looking for whether the industry was even talking about the fix. He searched a hundred and ten sessions tagged #ClimateWeekNYC for the word parametric. He found it zero times. None of what Swiss Re names as the cause of the gap is basis risk. Their own list is exposure growth, market access, pricing discipline, and data quality. It says nothing about the mismatch between what a trigger measures and what was actually lost. That mismatch is a separate problem, and it explains something narrower than the headline gap: not why the market is underinsured, but why the fastest, most transparent instrument available to close it, parametric insurance, has never scaled to take a meaningful bite out of the number. A parametric trigger fires on the cause: a rainfall total, a wind speed, a temperature threshold, measured by a satellite or a weather station some distance from the thing that was actually damaged. The loss occurs on the effect, the crop that failed, the reef that bleached, the livestock that died, and that is measured, if it is measured at all, on the ground, by a different instrument entirely, answering a different question. The two readings do not have to agree, and often they don't. A farmer can watch a field die below a rainfall threshold the policy never crossed. An insurer can watch a trigger fire over ground that, biologically, was never at risk. The payout and the loss are not the same event. They are two systems that were never built to talk to each other. The industry calls this basis risk and treats it as inherent, the cost of doing parametric business. It caps the product at specialist scale. It keeps it out of the institutional portfolios that need a payout to match a loss before they will underwrite at volume. The firm publishing the industry's own benchmark protection gap number is working from an incomplete picture, by its own account. Swiss Re's expected-loss modelling for that $424 billion figure excludeswildfireentirely, citing data limitations, on one of the fastest-growing perils in the portfolio. And their own case study from Italy in 2023 showed final hail loss estimates arrive at nearly triple the initial figure, once the data caught up with what had actually happened. This is not an accusation of concealment. It is a statement about the limits of the model everyone downstream is citing as settled. Look at how differently the two halves of this problem are measured. The atmospheric system is instrumented to exhaustion. Fourteen dynamical models and twelve statistical ones, run by institutions across four continents, in near-continuous agreement. NOAA's Climate Prediction Center puts the chance of a very strong El Niño this winter above90%. The World Meteorological Organization said on 3 September that the odds of it persisting through February are close to certain. Its own seasonal update forecasts the Pacific will run roughly3.6degrees above average at the peak. Paleoclimate researchers, reconstructing four centuries of ocean temperature from coral cores in the Galápagos, published findings this week placing the current event as the strongest in at least athousand years. That is what full instrumentation looks like. Every reading arguing with every other reading only over precision, never over whether the signal is real. Nothing measures what the ground beneath that signal can actually withstand, not to that standard, not at that scale. The atmosphere gets fourteen models. The substrate gets a rainfall proxy and an assumption. Close that gap and the trigger stops measuring the wrong thing. The payout moves from index to condition and starts matching the loss it was written against. That reframes what the protection gap actually is. It is not a distribution problem. The insurers already reach the exposed. It is not a capital problem. Capital is not what is short. It is not a translation problem. Everyone in that room can already read a rainfall chart. It is the absence of a single layer the industry never built: something that prices the asset's biological or physical condition, holds that price, and transfers it before the event, rather than settling for it after. That layer exists. Four pressures are converging in the same window, and none of them are waiting for the others to finish. The El Niño is the first. The trees are the second. The perennial monoculture belt planted across the world from the 1960s onward, coconut, palm, cacao, is ageing out of productive life together, because it was planted together, on the same clock, with no staggered succession built in. The people who know how to read that ageing are the third, and they are running out at the same rate as the trees. The global average age of a primary agricultural holder sits between 55 and 60. Under-35 engagement in farming worldwide sits near 10%. The knowledge that took four decades to accumulate, reading soil, reading rain, reading the gap between planting and harvest, does not transfer to a successor generation that left for the cities. The fourth is the one thing that could have bought some of that time back. Since the closure of theStrait of Hormuzin February 2026, the corridor carrying a large share of the world's nitrogen and phosphate fertiliser has been almost entirely shut, disrupting an estimated 38% of global nitrate-based supply and 20% of phosphate-based supply. The FAO has warned the effect lands inside thecrop cyclecurrently underway, tightening food supplies through the second half of 2026 and into 2027, because fertiliser applied late, or not applied at all, cannot be recovered later in the season. None of these four waits for the others. They are landing on the same ground, in the same eighteen months, and the insurance built to protect that ground still triggers on rainfall. For a household with no reserve to absorb a failed season, a season with no payout is not a setback, it is the end of the farm, and the family leaves for the city. The generation that would have inherited the knowledge, and the land, goes with them. What is left behind is not a temporary loss. It is ground nobody is left to run, insured by an instrument that was never built to see this coming. This piece stops at the mismatch. The next one describes what closes it. If you underwrite parametric risk, model catastrophe exposure, or price physical assets for a living, this is the one worth sitting with. The Natural Capital Trader publishes on biological time and institutional capital.