Goldman Sachs has estimated that Chinese gold buyers acquired 148 tonnes of gold between May and August 2026, more than double the 65 tonnes officially reported by the People’s Bank of China (PBoC). This 83-tonne difference represents a 128% gap between estimated purchases and the reserves disclosed by China’s central bank during the same period.
Discrepancy between official data and market estimates
The sharp divergence between market-based estimates and official reserve figures has focused global attention on China’s true gold demand. On October 19, 2026, several market reports highlighted this discrepancy, with analysts referencing robust trading activity linked to Chinese buyers in London’s over-the-counter (OTC) bullion market. In these OTC markets, trades are often not immediately reflected in formal reserve announcements.
Goldman Sachs’ model calculates inferred purchases using market flows and trading patterns, while PBoC figures reflect only formal changes in China’s reported gold reserves. As a result, these two measurements can differ due to methodology and timing.
Goldman Sachs’ estimates indicate that in August 2026 alone, China may have purchased 25 tonnes of gold in the London OTC market, which is 25% more than the 20 tonnes officially disclosed for the same month by the People’s Bank of China.
This gap does not directly prove that China secretly acquired the additional gold, but highlights a significant measurement difference in global bullion markets.
Mini dictionary: Over-the-counter (OTC) bullion market — A decentralized market where buyers and sellers trade precious metals such as gold directly without an exchange, resulting in transactions that may not be reported immediately in official statistics.
Monthly breakdown and evolving difference
A detailed review of the estimated and reported data shows that the widest gap occurred in May 2026. Goldman Sachs estimated Chinese demand at 48 tonnes that month, while the central bank’s disclosed reserve increase was around 10 tonnes, a 38-tonne difference. In June, estimates reached 40 tonnes versus 15 tonnes officially reported, while July saw 35 tonnes estimated against a 20-tonne addition. In August, the gap narrowed further, with estimated purchases of 25 tonnes and 20 tonnes reported by the PBoC.
This trend suggests that the monthly difference decreased over the period, falling from a 38-tonne gap in May to just five tonnes in August. However, the cumulative difference for the four-month period remains significant, driven mainly by higher inferred buying earlier in the year.
| Month | Goldman Sachs Estimate (tonnes) | PBoC Official Addition (tonnes) | Gap (tonnes) |
|---|---|---|---|
| May 2026 | 48 | 10 | 38 |
| June 2026 | 40 | 15 | 25 |
| July 2026 | 35 | 20 | 15 |
| August 2026 | 25 | 20 | 5 |
According to the World Gold Council, China’s official gold holdings reached 2,387 tonnes in August 2026 after an additional 20.2 tonnes were recognized by the PBoC. In September 2026, China reported a further 23 tonnes, extending its current gold buying streak to 23 consecutive months and raising its official reserves to 2,410 tonnes.
Global context: Central bank gold purchases
Goldman Sachs has also identified a similar measurement gap globally. For August 2026, the investment bank estimated total central bank gold buying at 79 tonnes, compared to just 39 tonnes officially reported by the World Gold Council. The estimated global figure is approximately 4.6 times the pre-2022 monthly average of 17 tonnes.
The World Gold Council uses public disclosures to compile central bank purchases, while Goldman Sachs relies on market activity to infer buying. Analysts warn that these figures are not additive and may overlap, meaning the inferred and officially reported numbers cannot simply be summed.
The World Gold Council’s 2026 survey found that 89% of central bank respondents expect their official gold holdings to increase in the upcoming year, while 74% predict a decline in the US dollar’s share of global reserves over five years. These survey results reflect forward-looking expectations, not confirmed trades.
The documented measurement gap underscores the complexity of tracking official and unofficial gold flows. While the figures point to significant ongoing purchases, they do not independently confirm undisclosed state-level acquisitions or provide definitive evidence of concealed reserves.
