Mongolia has three million people, two neighbours and one customer. In 2024 the world imported $14.8bn of goods from Mongolia. China bought $13.6bn of it: 92%.
A boom built on one buyer is a loan with no fixed repayment date.
Right now that customer is buying. Exports in the first eight months of 2026 were $14.4bn, 57% more than in the same months of 2025 and close to the total for any full year before. August was the biggest month in our data. The economy is growing at 7.7%.
Mongolia has been exactly here before. It is worth remembering how that ended.
No exporter in our data leans on a single buyer like Mongolia
2011: the fastest-growing economy on earth
In 2011 Mongolia grew by 17%, the fastest rate in the world, according to the World Bank. A vast copper and gold mine was being built in the Gobi, and China was buying every tonne of coal the country could truck south. In 2012 Mongolia borrowed $1.5bn on international markets at barely 5%.
Then China's steel industry slowed, coal prices halved and a dispute stalled the mine. By 2016 growth was 1.5%, reserves were down to $1.3bn and the central bank had raised interest rates to 15% to defend the currency. In February 2017 the IMF, China, Japan and South Korea put together a rescue worth $5.5bn, almost half the size of the economy.
Reserves are six times what they were at the last bust
What is different this time
Three things. Reserves are $8.0bn, six times the 2016 level. Government debt is 42% of GDP, down from 66% in 2017. And the export base has widened from one rock to two: in 2025 Mongolia's buyers recorded more copper concentrate than coal, the reverse of 2024, as the underground section of the Oyu Tolgoi mine in the Gobi ramped up.
Copper and coal do not move together. Coal follows Chinese steel. Copper follows power grids, electric cars and data centres. That is real diversification of product. It is not diversification of customer.
Copper overtook coal in 2025
What is the same
The boom is overheating the economy just as it did last time. Inflation was 12.9% in September and the policy rate is 12.5%. Total external debt, including what mining companies owe their parents, is $39bn against an economy of $25bn.
And the customer has shown before that it can stop. Between 2020 and 2022 China closed and restricted its border crossings for health reasons. Mongolia's reserves fell by more than a fifth in 2022 even with coal prices at a record.
Inflation is back near 13%, and so are interest rates
Finland, 1991
The best historical parallel for a one-customer economy is not a mining country. It is Finland.
Through the 1980s Finland sold ships, machinery and clothing to the Soviet Union under a barter arrangement that paid for its oil. At its peak the Soviet market took a quarter of Finnish exports. The trade was comfortable, protected from competition and politically useful to both sides.
In 1991 the customer ceased to exist. Finnish exports to the east fell by about 70% in a year. Combined with a banking crisis, it produced, by most accounts, the deepest recession any Western European country had seen since the 1930s: output fell by about 10% and unemployment rose from 3% to 17%.
Finland's exposure was a quarter. Mongolia's is nine-tenths. Finland also had ports. When Australia lost its Chinese coal sales in 2020, its ships simply sailed to India and Japan. Mongolia's coal leaves by truck, and the road goes to one place.
Exports in eight months of 2026 nearly match any full year before
Our expectation
We expect 2026 to be Mongolia's record export year by a wide margin, above $21bn. That needs $6.6bn in the last four months, or $1.65bn a month, and no month since March has been lower. We put it at three in four.
What would stop it is the same thing that always has: a decision in Beijing, or a sharp fall in coal demand from Chinese steel mills. The 2022 border episode is the model.
The more useful question is what Mongolia does with the windfall. In 2012 it borrowed against the boom. This time it is saving part of it. The currency, at about 3,600 to the dollar, is being managed rather than defended, and we would not bet against it while reserves are rising.
What to watch
A country cannot choose its neighbours. It can choose how much of a boom to keep.
- Monthly exports. Below $1.5bn for two months running would be the first warning.
- Chinese steel output. It still decides half the story.
- Reserves. The gap between $8bn and the $1.3bn of 2016 is Mongolia's real insurance.
- The 2027 budget. Booms end in the spending decisions made at the top.
Geoanalytic builds private data systems for advisory firms.
Visit the homepage →