The Country That Conquered Oil
China built an oil buffer just as its oil dependence began to shrink.
Grüezi!
Twelve weeks after Hormuz closed, the cleverest explanation of what China is up to, is also wrong.
The explanation comes from Izabella Kaminska at The Blind Spot. Chinese crude imports have collapsed since the US-Israeli strikes on Iran on 28 February and the closure of the Strait of Hormuz.
Yet Chinese refineries have only slowed modestly, transport still functions, and there are no visible signs of a national fuel shortage.
Kaminska’s answer is that China’s oil reserves may not really be usable. The barrels may have become collateral inside a hidden credit system: warehouse receipts pledged and re-pledged until the reserve itself became a kind of parallel financial network. Draw the oil, and the credit structure breaks.
It’s a smart theory, and Kaminska is one of the few writers brave enough to think aloud. But the public timeline doesn’t fit. Nor does the distribution of pain inside China itself.
Simpler? China built an enormous oil buffer over two decades, then entered the crisis just as its oil demand was beginning to plateau.
1. The puzzle
Since US-Israeli strikes on Iran in late February, Hormuz has been closed and China’s tanker-borne crude imports have fallen off a cliff.
They’re down from 11 million barrels per day to less than 6.5 million in May. That is nearly 40 per cent of China’s tanker-delivered crude taken out – and once you account for domestic production and pipeline crude – China’s economy is getting a fifth less oil than before the war.
But the scale of that collapse is barely visible.
China’s refineries have only cut back modestly. Sinopec, the largest state-owned major, is refining around 10 per cent less oil, and for the Shandong “teapots” – the independent refiners that handle 20 per cent of China’s crude – it’s a similar story.
There are no visible signs of shortages. Planes are still flying, trucks are still trucking, and rush hour traffic jams haven’t eased up.
China isn’t making up the shortfall drawing in more oil through Russia and Kazakhstan; those pipelines have hard capacity limits.
The difference has to come from somewhere.
Is China simply using up the gasoline, diesel and jet fuel that it has already refined and stored?
Or is crude oil being quietly drawn from China’s Strategic Petroleum Reserve, the state-held emergency stockpile? From the underground sites that no outside agency has ever counted?
China has three buffers to manage demand. First, what’s sitting in commercial tanks already refined: diesel, gasoline and jet fuel. Second, commercial crude: oil held by refiners and traders. Lastly, the Strategic Petroleum Reserve (SPR): the state’s emergency stockpile.
Kaminska’s theory depends on the SPR being financially locked. The argument I’m going to make is that the first two, plus pipeline crude and weaker oil demand, are buffering the shock – so China hasn’t had to test her proposition.
2. Over a barrel
Iza Kaminska is firmly in the follow-the-money school of analysis – and to explain what’s going on she has a thought experiment that taps into a very recognisable tradition.
In the 1770s, Genevan banks bought French annuities on the lives of carefully selected young girls, pooled those lives into packages of thirty, sliced the pools into shares, and sold the claims on. It was securitisation before the word existed: tradable paper resting on distant collateral.
Kaminska’s suggestion is that Chinese crude may have acquired a similar financial life. The warehouse receipts on the barrels in the Strategic Petroleum Reserve might have been pledged and re-pledged inside the country’s banking system over a decade.
They could have grown into a parallel-dollar clearing system: a private network of credit settled in oil-backed paper rather than actual dollars, with warehouse receipts standing in for the currency.
Paper claims behind credits, credits behind further credits, the whole thing resting on oil barrels that nobody will ever touch. Except, of course, in a crisis.
Drawing crude out of the reserve would force the pledges to settle. The settlements would unwind the credits. The result would be the equivalent of a run on the yuan.
If this were true then the reserves would be unusable, and China would have to manage an oil shortage without actually touching them.
If Kaminska is right, what would you expect to see?
Maybe an export ban to conserve whatever fuel is available? China imposed one on 4 March.
A domestic price cap to manage the imbalance? The National Development and Reform Commission, the NDRC – China’s macro-planning ministry – capped price increases at the pumps on 23 March.
A collapse in refining margins? By mid-May independent refining margins had collapsed deep into negative territory. Refiners are losing money on every tonne they process.
And then a state-led move to plug the financial hole. To keep the stock market steady, China has been buying its own state-owned enterprise shares for years through a group of government-backed funds – like Central Huijin – known collectively as “the national team.”
In the first half of 2026 it announced it was planning to sell roughly 90 per cent of those holdings. Kaminska sees this as the state cashing in stocks to cover the oil-shock shortfall.
Each prediction has come true. Kaminska’s case doesn’t rest on the size of the oil reserves; it rests on the argument that they can’t actually be used.
There is a precedent. In 2014, the same kind of analysis caught what Western commodity analysts had missed at the Chinese port of Qingdao, where traders had pledged the same copper and aluminium stockpiles to multiple lenders, sometimes ten times over. When that unwound it hit international banks causing billions in losses.
Nice theory. Is it true?
3. What the books say
China is importing less oil, building its crude reserves, refining slightly less, and running its economy more or less normally. That’s the entire, confusing picture.
Vortexa, the cargo-tracking service that estimates oil flows from satellite imagery, put Chinese crude stocks at 1.24 billion barrels by April – the highest ever, with over half a million barrels a day going into storage even as imports collapsed.
Some 360 million of those barrels sit in China’s Strategic Petroleum Reserve; the rest is in tanks belonging to refineries and traders.
If Kaminska were right, China would be in big trouble. Reserves it can’t use, a credit system about to fail, the state selling assets to paper over the cracks.
But China is sitting on crude reserves it isn’t drawing, because it doesn’t need to draw them. Individual commercial stocks are being redirected by the state, but the aggregate crude-stock position is not falling, it’s rising.
And a closer look at that share sell-off doesn’t help her argument.
Bloomberg estimates the “national team” will cut its ETF holdings by 90 per cent over the first half of 2026. The sell-off began in mid-January, with $123 billion sold in that month alone – the heaviest on record – including $67.5 billion over six trading sessions in late January. But US-Israeli strikes came on 28 February. Hormuz closed in early March.
The selling began six weeks before the strikes, seven weeks before Hormuz closed. Back then Brent crude was trading in the low $70s. Chinese refinery utilisation was at multi-year highs. Commercial crude reserves were building at the fastest rate on record. Whatever Central Huijin was doing in January, it wasn’t covering a shortfall from an oil shock six weeks in the future.
What might it have been doing instead? Bursting the bubble on tech speculation. In early 2025, the Hang Seng China Tech Index forward P/E reached 28, which is pretty expensive by historical standards. As the year progressed, Chinese investors were throwing even more money into AI and chip stocks traded in Shanghai and Shenzhen.
The “national team” sell-off was aimed at limiting their exuberance. It was concentrated in the Star 50 and CSI 300 trackers, which is where the tech bubble was biggest. And it has carried on into April and May – $30 billion since April says Bloomberg. This doesn’t look like a crisis-driven sell-off.
Highly unlikely that Central Huijin knew about US-Israeli planning six weeks in advance and started AI and tech ETFs in January.
Without this, the rest of Kaminska’s chain looks a lot like ordinary crisis management – export bans as supply management, price caps as inflation policy, margin collapses as the predictable side-effect.
4. Who takes the pain?
There is some pain in this particular China crisis, and it is being shared out by political rather than financial exposure.
The margin hit is going straight to the Shandong teapots. Independent refiners can’t buy crude directly the way the state-owned majors can; they get annual import quotas instead. On paper, the 2026 quota is the same as 2025 – 257 million tonnes. In practice, Beijing has tightened the tap. Quotas are released in batches, and this year’s first two batches were 30 per cent down on last year’s.
JLC, a Chinese commodity-pricing consultancy, put the teapots’ pre-crisis profits at a comfortable 240 yuan per tonne at the start of 2026. By March, however, as things hotted up, Beijing was pressuring them to keep processing crude at a loss, threatening quota cuts if they cut back. Losing quota is the worse threat: margins recover, but a teapot that loses its annual import allowance can take years to win it back, and one stripped of enough quota stops being a refinery at all. By early May the teapots were losing 500 to 600 yuan a tonne, and capacity had fallen from 55 to 50 per cent.
Beijing is willing to squeeze the teapots on margins and quotas, but it is not prepared to leave them exposed to US sanctions. Hengli Petrochemical, the second-largest independent, was listed by the US Treasury in April for buying Iranian crude. China’s Ministry of Commerce hit back with its first-ever Blocking Order – a legal measure under the 2021 anti-foreign-sanctions framework that bars Chinese firms from complying. China can discipline its private sector. No one else can.
State-owned majors have also taken losses. Sinopec disclosed an 830 million yuan first-quarter loss on liquefied natural gas imports after Qatar invoked force majeure – the contractual clause that lets a supplier suspend deliveries because of events outside its control – on its LNG exports. Sinopec president Zhao Dong said the company was cutting March refining by 5 per cent, with the option of cutting deeper in April and May.
In mid-March it asked Beijing for a 95-million-barrel allocation of commercial reserves to manage itself – roughly 40 days of its Middle East imports. Beijing refused. The interesting thing is what it didn’t refuse: the commercial reserves are being drawn at up to a million barrels a day through the central state channel. Sinopec wasn’t denied crude. It was denied the right to decide when to use it. The reserves are being used; but it’s the state, not the refiner, who decides when and how to use them.
PetroChina, the other state-owned major, is in better shape. It has 220 days of cover against its smaller seaborne exposure thanks to Russian pipeline crude and domestic production. At the end of March, its chairman, Dai Houliang, told reporters that only about 10 per cent of PetroChina’s total operating volume goes through Hormuz.
The difference between the two majors isn’t political. PetroChina’s source mix – Russian and Kazakh pipelines, plus a deep domestic production base in the Daqing, Changqing and Tarim oilfields – means it is pretty well insulated from the closure. Sinopec’s coastal refineries are designed to process heavier Middle Eastern oil grades that aren’t arriving. They can’t substitute easily. Politics allocated pain to China’s majors before its consumers; then supply networks decided which one took the worst of it.
China has also been running a deliberate substitution policy. Brazilian crude imports more than doubled year-on-year in March 2026 – 5 million tonnes against 2 million the year before. Russian seaborne deliveries rose 14 per cent.
Meanwhile Sinopec’s trading arm Unipec and Sinochem have been selling West African cargoes from their long-term contracts into Asia, where the lighter, ‘sweeter’ grades trade at a premium against the Hormuz-driven price spike.
And Indian refiners, who stepped back from Russian Urals under US pressure, have been buying.
The state trading desks are running parallel operations – disposing of wrong-spec contract crude on one side, buying medium-heavy substitutes on the other.
The winner – or the least inconvenienced party – is the Chinese consumer. Wholesale diesel is up 37 per cent and gasoline 26 per cent since the end of February. But at gas stations, the price cap has absorbed most of that before it reaches drivers.
This is what the state choosing who pays looks like.
Under Kaminska’s reading, you would expect pain distributed by financial exposure – institutions holding the most pledged collateral taking the deepest hits, with the state intervening to prevent contagion.
What’s actually happening is distribution by political priority: independents take a margin hit first, state-owned majors lose out in refining and on forecourts, households get protected.
If the problem was a credit chain, the state would be acting to prevent settlement events.
But if the problem is political – protecting consumer prices, keeping the majors in line, strong-arming the private sector – then there’s no hidden mechanism to discover.
5. The Chinese-language record
Chinese state media has been describing all of this in great detail since early March.
Just after Hormuz closed, on 4 March, Sina Finance said China was “preparing while others were panicking” (中东战火再燃，中国不慌). Between November 2025 and February 2026, China banked eighteen days worth of national consumption before the Strait closed.
Xiamen University’s Lin Boqiang sits on the National Energy Commission’s expert panel and advises the NDRC on energy prices. He told Huxiu on 13 March that China had enough crude to cover more than 100 days of national consumption. Also, Lin said, oil and gas accounted for less than a third of Chinese primary energy needs – as opposed to nearly two-thirds for the EU, almost three-quarters for the US.
A few days after, Cui Shoujun of Renmin University published “Chinese Resilience” (中国韧性) on the Foreign Ministry’s “Xi Jinping Diplomatic Thought” channel – the platform Beijing uses to share approved messaging to Party cadres and foreign analysts.
China, said Cui, had spent the past decade readying itself for a crisis like this: it had diversified its suppliers; found alternative routes (Kazakh and Russian pipelines, China-Iran rail); built up its strategic reserves; and developed options to pay for oil in renminbi rather than dollars.
Well, Chinese experts would say that, Kaminska might object. This is just state-choreographed expectation management.
But if a parallel-dollar mechanism existed, it would almost certainly leak out somewhere beyond state media – in shadow-banking commentary, warehouse-receipt stories, investigations of commodity financing in Caixin, the financial magazine that produces the most clear-eyed coverage of China’s banking sector.
This stuff did surface in Chinese-language financial press during the 2014 Qingdao episode that Kaminska herself reported on. It’s not circulating now. Caixin’s coverage of the refining sector is the usual fare: margins, quota allocation, and capacity rationalisation.
6. The blind spot
How much crude China is actually sitting on is anyone’s guess. Vortexa’s 1.24 billion barrels is one estimate. The US Energy Information Administration’s reckons it’s around 1.4 billion. Kpler-sourced numbers run above 1.5 billion.
That’s a 300-million-barrel gap between those paid to know – enough oil to run the country for the best part of a month. Lin Boqiang’s 100-day figure sits at the upper bound – 100 days of Chinese consumption is around 1.4 billion barrels.
Which means nobody outside the system can model China’s energy buffer with any confidence.
But the buffer isn’t being drawn down yet. The diesel, gasoline and jet fuel already refined and sitting in commercial tanks is filling the gap, with pipeline crude from Russia and Kazakhstan keeping the refineries running. The crude reserves are the next line of defence, not the current one.
And even if they were being drawn, they’d last a long time. The commercial reserves alone – the roughly 880 million barrels held in tanks by refineries and traders, separate from the strategic reserve – would run for the best part of a year at realistic draw rates.
The second unknown is harder. How much oil does China still need?
China is in the middle of the fastest oil-demand transition any major economy has attempted. Battery and hybrid cars were 48 per cent of new sales in 2025 by the official Chinese auto industry figures – and that crossed half on a monthly basis from October onwards. The number is still rising.
In 2025, electric heavy trucks were 22 per cent of new sales in the first half of the year and 28 per cent by August – taking out an estimated million barrels a day of diesel demand.
Sinopec itself has said Chinese gasoline and diesel demand have peaked. CNPC’s research institute has total Chinese oil demand peaking in 2025. The IEA brought forward its China-peak forecast by two years to 2027.
The transport-fuel buffer is deep. The petrochemical one, less so. Naphtha and LPG imports through Hormuz feed China’s plastics and synthetic fibres industries, and if the Strait stays closed those inputs will start to squeeze factories by autumn.
Still, on garage forecourts, none of this looks like an economy in distress. If it’s a stress-test then it looks like one the country happened to be unusually well-prepared for.
When Hormuz reopens, Beijing will at some point start buying crude on a normal scale again.
But, the question for markets is what will the new ‘normal’ look like?
7. The elegance trap
Follow-the-money writers read the world through pledged collateral and credit chains. It’s an approach that’s led to some real scoops, including the 2014 Qingdao scandal where the same copper and aluminium were pledged to ten different lenders. Kaminska called that one.
The trouble is that the same instincts can read a crisis into a coincidence. The “national team” selling that Kaminska reads as cover for a crisis started six weeks before the crisis. Correlation but no causation.
Hidden chains, parallel-dollar clearing, state liquidations to cover shortfalls – these are all better stories than refinery quotas, EV adoption, and price-capping bureaucrats. They’re also not what’s happening.
The elegant story is that China’s reserves are unusable because they are trapped inside a hidden credit machine. The uglier story is that China built large oil stocks, controls who takes the hit in a crisis, and is using refined products, commercial crude, pipeline supply and weaker demand to get through a shock without a visible household crisis.
The evidence fits the ugly story.
China does still need its oil reserves. Hormuz, sanctions, insurance shocks and war risk haven’t disappeared. But it no longer needs those reserves in quite the way it once did.
The state’s China Energy Outlook 2025-2060 has total oil consumption falling about 80 per cent from peak by 2060 – chemical requirements included.
Twentieth century geopolitics was built around oil. In the first quarter of this century, Beijing has hedged its geopolitical risk by stockpiling oil. But also through electrifying its economy.
The Hormuz crisis has arrived just as that economy is starting to become less oil-dependent.
That’s the real blind spot. The market is trying to estimate how much oil China has. It may be underestimating how much less oil China will need.
Thanks for reading,
Bis bald!
Adrian
PS As I was writing Michal Meidan dropped this excellent paper.