China Could Crash The Price Of Oil

Oil tank farm along NJ Turnpike in Linden, New Jersey October, 1979. (Photo by Jack Mecca/Getty Images)

Jack Mecca

The internet is awash in skepticism about the supposed looming oil glut as commentators note that, despite the IEA’s implied inventory build of 800 million barrels this year and 1200 million barrels next year, most of that has not shown up in the data. The figure below shows the implied inventory build calculated using the IEA’s global supply minus global demand and assuming OPEC production remains constant from June 2025 to end 2026. The implication is of a massive glut and inventory build, already putting pressure on prices.

Source: Supply minus demand from IEA Oil Market Report; OPEC production assumed flat from June 2025.

Responses to this have included trashing the IEA as either incompetent or dishonest and some are also arguing that the failure of inventories to grow implies a much tighter market now and in the coming months than the IEA is forecasting. Certainly, the IEA has often underestimated demand outside of the IEA member countries, where data ranges from lagged to poor; if that is the case now, the market balance is much less threatening that projected.

Some of the criticisms are off-base. The IEA’s forecasts are biased in the sense that they are produced by humans and humans are biased. The same can be true of market reports from both the EIA and OPEC and the punditosphere. And while there might be institutional biases, these are not straightforward: the IEA has often warned of coming price spikes, yet now we have the IEA forecast implying an exceptionally weak market, as if their bias was somehow reversed.

Without doubt, inventories are not behaving as they should. The figure below shows inventory changes compared to the implied balance (which is supply minus demand). The numbers for 2024 appear reasonably well matched, with the inventory changes close to what would be expected. That changes this year, when there should have been an enormous inventory build, about 500 million barrels of 1.5 mb/d in the first half, versus ‘observed’ inventory builds of only 0.5 mb/d.

Source: IEA Oil Market Report; Industry and government stocks for OECD countries.

So where’s the missing oil? Inventories outside the IEA countries are not included in the data above and, given that half of oil consumption is outside the OECD, it would be expected that inventory builds in the OECD, shown in the figure above, would only be about half of the total. More important, the IEA reports that from April to August this year, Chinese inventories grew by 110 million barrels, or over 700 tb/d. (The EIA estimated the number at 900 tb/d from January to August. Expanding strategic oil stocks in China support crude oil prices – U.S. Energy Information Administration (EIA) This easily explains the discrepancy without needing to assume demand is 1 mb/d underestimated. It might be, but then the inventory picture would look much tighter than the current data shows.

In terms of the market going forward, then, the level of confidence in the IEA’s projected market balance should be that much greater. And even if demand estimates are off by 1 mb/d or so, the market surplus still looks to be over 2 mb/d for 2026. If one assumed that half of that oil went into OECD inventories, as shown in the figure below, then it is hard to avoid expectations that the market will be under significant pressure in the months ahead.

Data from IEA. Forecast from July 2025 based on assumption that half of difference between supply and demand in IEA forecast goes into OECD commercial inventories.

The nature of this year’s inventory build, apparently heavily concentrated in Chinese government inventories, is important because behavior varies enormously between government and commercial inventory holders. Commercial inventory holders will sometimes increase their stocks if they expect prices to go up and sell them off if they think future prices will be lower, which exacerbates the price cycle. For their part, government inventory holders tend to buy and hold as a hedge against future supply disruptions and without regard for market developments.

China’s government stock-buying has three potential motivations: their desired inventory levels are determined by imports which had been increasing; the possibility that Russian oil supplies will be affected by sanctions leads them to add to holdings as a hedge; and the possibility that the government fears some political dispute will lead to efforts to embargo their oil purchases. The last could reflect either Chinese government plans to take some aggressive actions that it fears will trigger an embargo or simply an assessment that U.S. policies are so unpredictable that an increased hedge against a possible embargo is needed.

Chinese decision making regarding the desired level of oil stocks is obviously not transparent; changes are observed but with a lag, but what drives them to buy or not is unknown. Since we don’t know either precisely why the government has been adding to its strategic stocks or what their target level is, the future purchases (or sales) cannot be predicted. The very real possibility that purchases might abruptly cease means the market balance could shift quickly and severely.

If the global economy is strong over the coming year and sanctions against Iran, Russia and Venezuela are tightened, then the market balance could prove to be much tighter than the IEA projection. A surplus of 2 mb/d is quite possible as an annual average, but the first half could still see an inventory build around that amount even in a bullish price scenario. That wouldn’t be too difficult for markets to accommodate—assuming China continues to buy 1 mb/d for its strategic stocks. But cessation of those purchases by China would threaten to turn an inventory surge into a flood and OPEC+ will have to step in again to prevent a price collapse.