In June 2023, northern Vietnam ran out of water before it ran out of demand. A severe drought and a record heatwave drew down the reservoirs feeding the north's largest hydropower plants, and eleven of them, including Son La, the biggest in the country, fell to levels too low to run. Around 5,000 megawatts left the grid almost at once.[1] Factories rationed power, some with little warning, and the affected sites included suppliers to Samsung, Foxconn, and Apple. The World Bank later put the cost of the episode at about 1.4 billion dollars, roughly 0.3 percent of GDP.[2]
None of it was a construction fault or an operating error. The machines worked. The water simply was not there.
For anyone taking equity in hydropower in the Lower Mekong, that distinction is the whole point. A hydropower asset is usually described by what it owns: the dam, the turbines, the concession, the long-term offtake. The more useful question for an equity investor is what it does not own. In this basin, the answer is the one input that determines whether the asset performs at all.
The asset you buy is not the river you depend on
Laos has spent two decades positioning itself as the battery of Southeast Asia. Around 80 percent of the electricity it generates is exported, mostly to Thailand and Vietnam, and those exports account for close to a third of the country's exports by value.[3] In 2024, power sales brought in more than 2.6 billion dollars, close to a quarter of total export earnings, and hydropower supplied roughly three quarters of domestic generation.[4] The largest mainstream project, Xayaburi, sends about 95 percent of its output to Thailand under long-term contract.[5] This is real, contracted, hard-currency infrastructure revenue, and it is exactly the profile that draws equity into the sector.
The difficulty is that the equity claim sits on top of a resource the concession does not control. An investor owns the plant and the contract. It does not own the rainfall, the snowmelt in the headwaters, or the reservoir decisions made upstream that determine how much of the river actually reaches the turbines. In a normal year this is invisible, because the water arrives and the model holds. In a dry year it becomes the only thing that matters, and it is governed by forces that are neither in the financial model nor within the investor's reach.
Drought is only half the risk; the other half is operated upstream
The first force is climate, and it is the one the market discusses. A strong El Niño weakens the monsoon that fills the basin, and the Lower Mekong has been through severe dry sequences in the last decade, most acutely in 2019 to 2020. The second force is less discussed and, for an equity holder, more uncomfortable, because it is not weather at all. It is policy.
The upper third of the Mekong sits inside China, where a cascade of large dams now regulates the flow before it crosses into Southeast Asia. In the dry season, as much as 70 percent of the water reaching the first downstream gauge in northern Thailand originates from that stretch of the river.[6] During the 2019 drought, satellite monitoring found that upstream dams held back enough water to suppress the annual monsoon rise at that gauge almost entirely, even though the Chinese section of the basin had received near-average to above-average rainfall that year.[7][8] In July 2021, a restriction at the Jinghong dam dropped downstream levels by half a metre in ten hours, and by mid-August one Thai district was missing about 35 percent of its natural flow.[9]
China disputes that reading, and points to dry-season releases and flood moderation as benefits to downstream neighbours. For an equity investor the argument over intent is beside the point. What matters is structural. A material driver of your asset's output is an upstream operator you cannot influence, in a basin where water has at times been treated as a sovereign resource rather than a shared one, and where the timing of releases has tracked the state of regional relations.[10] El Niño sets the baseline for a dry year. Upstream operation sets the tail. Both sit outside the concession, and the river supports around 60 million people downstream, which means the politics around it will never be simple.[11]
The offtake moves price risk, not water risk
The instinct in infrastructure is to point to the power purchase agreement and treat the asset as an annuity. The tariff is fixed, the offtaker is a state utility, and the revenue looks like a coupon. That reading confuses two different risks. A PPA transfers price risk. It does not transfer water risk. Contracted energy in these structures still has to be generated, and generation depends on inflow. When the water does not arrive, the contracted volume is not delivered, and the shortfall does not disappear. It lands somewhere.
Where it lands is the part equity should sit up for. Equity is the residual claimant. It is paid after operating costs and any senior obligations, from whatever cash the asset actually produces. A dry year does not shave a little off the top. It removes the marginal output the distribution was counting on, and it does so in the years the model treated as ordinary. Vietnam in 2023 is the cautionary version at national scale: a hydro-heavy northern grid, a drought, eleven plants offline, and a shortfall that became a 1.4 billion dollar macroeconomic event within weeks.[1][2] Even Laos, the exporter, is not immune to its own hydrology. In 2024 it spent more than 177 million dollars importing electricity during the dry season, because when the reservoirs are low the surplus it sells in the wet season is simply not there to sell.[4][12]
State it plainly, and Mekong hydro equity is a claim on the residual output of a variable, weather-dependent, partly foreign-governed input, sold to buyers under contracts that assume the input shows up. In a wet decade that is a fine business. Across a full climate cycle it is a different risk than the annuity it resembles.
Correlation is the risk you cannot diversify inside the basin
The final problem is that this risk does not diversify the way a portfolio manager would hope. Two hydropower assets on the same river system are not two independent bets. They share the same monsoon, the same El Niño signal, and in the upper basin the same upstream operator. Adding a second Mekong plant to a book does not spread the risk. It concentrates a single factor that happens to be driven by weather and foreign policy at the same time. That is the definition of an exposure that cannot be diversified from inside the basin.
The cycle then works against you at both ends. The same warming that deepens the droughts also raises the ceiling on the wet years that follow, and a strong El Niño historically discharges into a wetter phase within a year or two. An asset drawn down hard through a dry sequence can face the opposite tail next, when saturated catchments force spill and the water arrives as a hazard rather than a resource. For equity, that is variance in both directions around a return line the model drew as a smooth average.
What this means for investors
For a mid-sized infrastructure investor taking equity in Mekong hydro, the implication is not to avoid the sector. It is to stop underwriting it as a contracted annuity and price it as what it is.
Start with the entry return. Transboundary water risk belongs in the discount rate and the entry yield, not in a footnote, because it is a genuine and uncompensated source of variance in distributable cash. Then read the offtake for what it actually allocates. The question is not whether there is a PPA, but how much of the contracted revenue is genuinely firm regardless of inflow, and how much is volume that only exists if the water does. Deemed-generation and water-availability provisions are where that line is drawn, and they deserve as much diligence as the tariff itself.
Be honest about your position relative to the upstream operator. You are a price-taker on water, and no amount of downstream structuring changes that. Where the exposure is unavoidable, the mitigants sit at the equity level, through political risk cover and, often, through co-investment alongside a development finance institution whose presence changes the calculus around water sharing and offtake reliability. Finally, build the portfolio so that Mekong hydro is a position, not a theme. Pair it with assets whose output is uncorrelated to the same monsoon, consider hybridising reservoirs with floating solar to stretch a given volume of water further, and let storage carry the firming the river cannot promise.
Closing view
The strengthening El Niño now expected to peak in the 2026 to 2027 winter will test this basin again, and the official forecast already puts the odds of a very strong event well above even.[13] When it does, the assets that hold up will not be the ones with the strongest contracts on paper. They will be the ones whose owners priced the water they do not own before the dry season, rather than discovering the exposure in a distribution they can no longer pay.
The deeper point runs past the Mekong. As the grid decarbonises, more of the region's power depends on inputs that are variable, and some of them cross borders no concession controls. Hydropower is simply where that reality is most visible today. The equity investors who do well in it will be the ones who treat the river as what it is: not a fixed asset, but a shared and contested flow, priced accordingly.
References
- Asia Pacific Foundation of Canada. (2023, June 21). Northern Vietnam plunged into darkness as power crisis prompts investor concerns. https://www.asiapacific.ca/publication/northern-vietnam-plunged-darkness-power-crisis-prompts
- World Bank. (2023, August). Taking stock: Vietnam economic update, as reported in Power crisis cost Vietnam $1.4 bn: World Bank, Agence France-Presse. https://www.energy-daily.com/reports/Power_crisis_cost_Vietnam_14_bn_World_Bank_999.html
- Foreign Policy Research Institute. (2024, April). The battery of Southeast Asia: Challenges to building a regional transmission grid. https://www.fpri.org/article/2024/04/the-battery-of-southeast-asia-challenges-to-building-a-regional-transmission-grid/
- The China-Global South Project. (2026, February 3). As China builds the grid, Laos's power ambitions enter a new phase. https://chinaglobalsouth.com/analysis/laos-battery-of-southeast-asia-china-power-grid-2026/
- Andritz. (n.d.). Laos: The battery of Southeast Asia. Hydro News. https://www.andritz.com/hydro-en/hydronews/hydro-news-asia/laos
- Eco-Business. (2021, March 18). New platform aims to reveal dam and climate impacts on the Mekong. https://www.eco-business.com/news/new-platform-aims-to-reveal-dam-and-climate-impacts-on-the-mekong/
- Stimson Center. (2020, April). New evidence: How China turned off the tap on the Mekong River. https://www.stimson.org/2020/new-evidence-how-china-turned-off-the-mekong-tap/
- Al Jazeera. (2020, April 25). China's dams exacerbated extreme drought in lower Mekong: Study. https://www.aljazeera.com/news/2020/4/25/chinas-dams-exacerbated-extreme-drought-in-lower-mekong
- Dialogue Earth. (2022). Wet-season dam operations hit Mekong ecology and communities. https://dialogue.earth/en/water/hydropeaking-mekong-wet-season-regulation-hits-ecology-communities/
- Voice of America. (2021, January 29). China's diversion of upstream Mekong flows seen drying up Southeast Asia. https://www.voanews.com/a/east-asia-pacific_chinas-diversion-upstream-mekong-flows-seen-drying-southeast-asia/6201366.html
- Mongabay. (2020, April 30). China held water back from drought-stricken Mekong countries, report says. https://news.mongabay.com/2020/04/china-held-water-back-from-drought-stricken-mekong-countries-report-says/
- ASEAN+3 Macroeconomic Research Office. (2023, December 13). Lao PDR's electricity sector: Risk or opportunity? https://amro-asia.org/lao-pdrs-electricity-sector-risk-or-opportunity/
- National Oceanic and Atmospheric Administration. (2026, June 11). El Niño forms, expected to strengthen, say NOAA forecasters. https://www.noaa.gov/news-release/el-nino-forms-expected-to-strengthen-say-noaa-forecasters