Nepal's power-export boom runs on a river it does not control

On the morning of August 26th a wall of water came down the Bhote Kosi, the river that marks much of the Nepal-China border. The sky was clear and there had been no local rain; the surge simply arrived, and in minutes it folded villages, bridges and a customs yard into a brown torrent. By the following morning the confirmed death toll stood above twenty, with hundreds more missing — Nepal's tourism authority counted 384 foreign tourists among them — and the national grid had lost 431.1 megawatts (MW) of generation, roughly a tenth of the country's installed capacity, at the peak of the export season. Disaster relief is the government's problem. What should interest an investor is the machine the flood just switched off: a buildout that has turned park-bench-clean hydropower into one of Asia's fastest-growing export trades, on the assumption of a river Nepal does not control.

The buildout is real and still accelerating. Nepal added more than 500 MW of generation in a single fiscal year, pushing installed capacity past 4,200 MW, and in the year that ended in July it earned a record 29.3bn Nepali rupees (about $190m) selling electricity to India and Bangladesh — roughly two-thirds more than the year before. The stated ambition is ten gigawatts of exports within a decade, sold to a buyer, India, whose own solar-and-storage buildout may not keep that market open forever. The corridor underneath the ambition is the one that flooded.

The Bhote Kosi descends out of Tibet through Rasuwa district, and the valley carries an unusually dense string of power plants. The Energy Ministry listed the 216 MW Upper Trishuli-1 — a flagship plant under construction, backed by Korean investors and the Asian Infrastructure Investment Bank — among the projects affected, along with a 25 MW solar farm. The Nepal Electricity Authority spoke of twelve affected projects, with six plants and a solar farm directly damaged and one reportedly washed away wholesale; the flood also severed a national transmission line.

A run-of-river plant lets the current turn its turbines; water is its only raw material and, in the model, it is free and steady. So the immediate blow to the sellers of power is a season of thinner export revenue, not an existential one. The existential problem, for anyone who owns or finances the sector, is the cause and the schedule.

What fueled the flood is still being established, and official explanations split. The district's chief administrator noted there had been no rain in the district headquarters; reports elsewhere blamed an ice-and-rock avalanche or a glacial lake outburst high in the watershed, disagreeing on whether the trigger lay on the Tibetan side or on Nepal's own slopes. The exact source remained under investigation. It would be the same family of event that struck the same stretch of river in July 2025, when a supraglacial lake in Tibet's Gyirong county drained without warning, killing at least nine people, washing away the Miteri bridge and damaging the Rasuwagadhi dry port. It would echo 2016, when a burst from a lake called Gongbatongshaco wrecked Nepal's other road to China and damaged the Upper Bhotekoshi power station. Climate researchers say the region's ice loss has roughly doubled since 2000, and the rubble-dammed lakes that store the danger are growing. A decade ago this was a once-a-generation event for one corridor. It has now happened twice within fourteen months.

The deeper problem is not the flood's force but its geography. The lakes, and the weather that will burst them, sit in Tibet on China's side of the divide. The dams, the roads, the dry port and the towns sit in Nepal. The party best placed to reduce the risk — China, which holds the monitoring data and could drain dangerous lakes or share early warnings — bears almost none of the downstream cost, while the party that pays for every recurrence cannot even see the source. The corridor gets rebuilt rather than redesigned because it is not optional: after the 2015 earthquake destroyed Nepal's old road to China at Kodari, Rasuwagadhi became the country's principal overland artery to its northern neighbour. The flood severed it on both sides; Chinese state media reported severe casualties on the Tibetan side too.

For investors, the telling thing is what the financial model does with this. Floods are classified as force majeure — one-off events that no contract prices — and so they are left where they stand: insurance for Himalayan megaprojects is thin, and the uninsured remainder falls to Nepal's treasury and its concessional borrowing. The recurrence interval exposes the accounting. Three catastrophic bursts on the transboundary system since 2016, with the gap shrinking from nine years to fourteen months, is not a run of bad luck; it is a cost with a schedule. Nothing in the sector's power-purchase agreements, financing costs or share prices yet charges for that schedule. And the flood is not even the only squeeze on the machine's economics: India, the biggest buyer of Nepali surplus, is adding solar and storage so quickly that analysts already warn Nepal cannot count on the market lasting. The export boom faces a tail on one side and an exit door on the other.

Who pays, then? First the Nepal Electricity Authority and the treasury — repairing plants, roads and lines — then the sponsors and lenders of the works in the valley, up to the development banks funding Upper Trishuli-1. The builders and equipment suppliers of Himalayan hydro, Korean and European here, Chinese state groups on the megaprojects, are diversified giants for whom Nepal is a rounding error. Indian buyers will shrug and build their own. For the American retail investor the honest conclusion is that there is no clean way to own this trade: Nepali hydropower developers list in Kathmandu, not New York, and the exposure otherwise hides, too dilute to matter, inside broad emerging-market and infrastructure funds. That scarcity is itself the finding. The risk is not priced, traded or hedged anywhere an ordinary investor can reach; it sits, uncharged, inside the expected returns of projects and the borrowing costs of a small country.

Watch instead for the moment it starts to be charged. It will be visible, if it ever comes, in the terms of the next dam: whether new project debt demands flood-proofing and insurance, whether Kathmandu ties reconstruction money to a transboundary early-warning and lake-drawdown arrangement with Beijing, whether lenders ask what happens to their collateral when the river that feeds it is not theirs to control. Until then, a growth story with a single physical corridor and a threat on the wrong side of a border will keep delivering its "one-offs" on schedule, and investors who extrapolate the good years will keep paying for them in forms they do not yet recognise. The Bhote Kosi has now offered its warning three times in a decade. The tragedy for Nepal is that it has no choice but to keep building. The lesson for an investor is to look for the price of a river it does not control.