On 11 June 2026, NOAA stopped hedging its language. The agency moved from an El Niño Watch to an El Niño Advisory, the difference between "likely" and "here," and put the odds of a very strong event peaking next winter at 63%, a level that would rank among the largest in the record since 1950.[1]
Most of the coverage that followed went where it usually goes: cocoa, coffee, palm oil, the price of breakfast.[2] That is the familiar El Niño story. The one that matters more for anyone deploying capital into Asia-Pacific power is quieter. A grid built on renewables does not burn fuel. It runs on weather. And the climate pattern now strengthening is the single largest source of year-to-year swing in that weather across the region.
This is a good moment to fix a mental model that most energy investors are carrying around half-built. The transition did not remove weather risk from the system. It moved it, and it changed the risk in three specific ways. Each one breaks an assumption that standard project finance still relies on. This piece sets out the frame. The articles that follow will test it asset by asset.
One: it moved to the revenue line
Under a thermal grid, weather barely touched the economics. The swing variable was fuel price, which markets hedge fluently with futures, swaps, and long-dated contracts. On a solar, wind, or hydro asset there is no fuel. The swing variable is the resource itself, how much sun, wind, and water arrive, and when. There is no deep, liquid market in "monsoon strength." That risk does not sit in a cost you can lay off on a screen. It sits in revenue you may simply never earn.
India shows the mechanism plainly. The country concentrates more than half its annual wind generation in the June to September monsoon. When the monsoon underperformed in 2020, peak-season wind output came in 24% below the prior year, and the western region fell 29%.[3] No input got more expensive. The output was not there. For a project financed against an average year, that is not a soft quarter. It is a debt-service problem.
Two: it made the risk correlated
A book holding wind, hydro, and solar feels diversified. El Niño can collapse that diversification in a single season, because the technologies share a weather driver.
The El Niño signature in South Asia is a weakened monsoon, and a weak monsoon does two things at once. It slows the winds the peninsular wind fleet depends on, and it cuts the rainfall that fills hydro reservoirs. In 2020, India still had enough water in its dams to ramp hydro and cover the wind shortfall. In a strong El Niño, that buffer is exactly what disappears. Drought linked to the 2023 El Niño cut India's hydropower generation by 8% year on year in the first half of 2024.[4] Put a weak-monsoon wind year and a drought-driven hydro year together, which a strong El Niño can do, and the two assets fall at the same time, for the same reason. Solar offers less cover than people assume here too: clearer skies lift the raw resource, but hotter modules and haze routinely erode the gain. Correlation, not exposure on its own, is what turns a manageable risk into a covenant breach.
Three: it made the risk serial, not random
Standard project finance leans on a typical weather year and a smoothed historical average. That quietly assumes bad weather is independent and short-lived, a cloudy week here, a calm month there, averaging out over the life of the asset. El Niño does not behave that way. It arrives as a sustained block, and its swings can run for years, as the multi-year La Niña that ended in 2023 showed from the other direction.
Tasmania is the cautionary case. The state runs largely on hydropower. An El Niño-linked spring drought in 2015 produced the worst inflows on record, and by late April 2016 the hydro system had fallen to around 12.5% of storage. The undersea cable that could have imported power had failed, and the state installed 200 MW of temporary diesel generation to keep the lights on.[5] The cause was not mismanagement. It was a sustained dry block of the kind a typical-year model is built to average away, and therefore cannot see coming.
What follows from this
Put the three together and the risk on a renewables grid is on the revenue line, correlated across the portfolio, and lumpy in time. That is close to the worst combination an underwriter can face, and it is precisely the combination a conventional model is designed to miss. Two consequences follow, and they preview the rest of this series.
The first is that the downside cases are too optimistic. A P90, the figure lenders treat as a conservative floor, is only as conservative as the weather record behind it. Built on smoothed averages, it understates the depth of a multi-year resource drought. One study of onshore European wind projects found them delivering a median of 8.9% below their P50 estimate, with only about one in seven matching or beating expectations.[6] If the central case runs high, the so-called conservative case is not conservative enough.
The second is that storage is the structural answer, not an accessory. Because storage earns from volatility rather than suffering it, its value rises exactly when the weather turns difficult. Australia's market gives the cleanest evidence. As volatility climbed, the average price batteries set when they were the marginal supplier nearly doubled in a year, from about A$245 to about A$478 per megawatt hour.[7] By early 2026, batteries, combining charging and discharging, were setting the market price in 32% of dispatch intervals, more often than any other technology, having overtaken hydro.[8] The asset most models still treat as a bolt-on is the closest thing the new grid has to a weather hedge.
None of this is an argument against the energy transition. It is an argument for financing it with the right model. The old one treated weather as an operating detail, something the control room handled day to day while the financial risk lived in the fuel price. On the grid we are now building, weather is the financial risk. It has moved from the night shift to the investment committee, and the tools have not all made the same move. A warmer baseline only steepens the curve, deepening droughts and loading the wet years toward heavier extremes, which raises the variance the model has to carry.
Over the coming weeks I will take this apart piece by piece and market by market: why the underwriting math runs optimistic, why hydropower in the Mekong is really a sovereign-risk question, why the good-for-solar assumption misleads across Southeast Asia, and why storage is quietly becoming the hedge of record. The starting point is the simplest part. The grid runs on weather now, the weather is about to make a strong move, and the single update worth making today is to stop pricing it as background noise.
Sources
Primary sources and market references cited above.