China Bought a Record Amount of Gold in 2026. The Price Fell Anyway.

Two facts from this year sit awkwardly together.

In the first eight months of 2026, China imported 1,141 tonnes of gold, worth roughly USD 158.8 billion. That is more gold than China bought in the whole of 2025, when imports totalled about USD 96.5 billion. It is the fastest import pace on record going back to at least 2017.

Over the same period, gold fell from USD 5,589 an ounce in late January to around USD 4,174 in mid October. A decline of roughly 25 percent from the peak.

The largest physical buyer in the world bought more gold than ever recorded, and the price went down.

Most commentary on this subject assumes the opposite relationship. Understanding why it does not hold is the single most useful thing you can know about gold right now.

The scale of what China has been doing

Set the price aside for a moment, because the demand numbers are remarkable on their own.

Imports. 1,141 tonnes through August, per customs data. Hold that monthly average through December and the year lands near 1,710 tonnes. Worth noting that ANZ puts the same eight-month figure closer to 1,000 tonnes, a gap that comes down to data sources and method rather than anyone being wrong. Either number is a record.

The central bank. The People's Bank of China has now added gold for 22 consecutive months. August's purchase of 20.2 tonnes was its largest since October 2023, and the bank has bought in double digits every month since May. Official holdings sit at roughly 2,387 tonnes. The valuation of those reserves rose from USD 306.35 billion at the end of July to USD 350.08 billion at the end of August.

Retail and funds. Chinese gold ETFs added about 44 tonnes through August, up 18 percent on the year, at a time when global ETF flows were close to flat.

The wider picture. Global central bank purchases are running at around 130 tonnes year to date against roughly 160 tonnes in the same period last year. So official buying overall has cooled, and China has been taking a larger share of a smaller pot. When analysts say China is leading sovereign gold demand in 2026, that is what they mean, and it is accurate.

Why China is buying

Four things are driving it, and none of them are about the gold price.

There is nowhere else to put the money. Chinese households face a narrower set of investment options than savers in most developed markets. Deposit rates are low, the property market has not recovered its role as the default store of household wealth, and capital controls limit offshore alternatives. Gold is one of the few assets available that is liquid, tangible and outside the banking system.

The mainland price has been running above the world price. When Shanghai trades at a premium to the international benchmark, importing gold becomes profitable for wholesalers. Strong local investment demand pushed that premium open, and imports followed the arbitrage. This is the mechanism, and it is worth understanding: the imports are a consequence of local demand, not an independent act of policy.

A strong yuan opened the door. The currency has held up well through 2026, and regulators responded by approving more generous import quotas. As Zijie Wu of Jinrui Futures put it, the strong yuan created favourable conditions for gold imports. Quotas are the tap, and the tap was turned up.

The central bank set the tone. Twenty two straight months of official buying is a signal retail investors read clearly. The PBoC's own purchases are modest next to import volumes, around 80 tonnes in eight months, but their influence on domestic sentiment is considerably larger than their tonnage.

Notice what is absent from that list. None of it is a bet that gold will be higher next quarter. China's buying is structural and largely price insensitive, which is exactly why it has continued through a falling market.

So why did the price fall

Because the gold price is not set where the gold is bought.

Price discovery happens in London and New York, in futures and ETFs, and it responds to two things above all others: real interest rates and the dollar. Both turned against gold in 2026.

The Federal Reserve raised rates by 25 basis points to a 3.75 to 4.00 percent target, its first increase in over three years. Bond yields rose with it. The dollar index climbed to around 102, its strongest in roughly eighteen months. Every one of those moves raises the opportunity cost of holding an asset that pays no income, and makes gold more expensive for anyone buying it in another currency.

There was also a crowded trade to unwind. Gold rose a long way into January 2026. A move of that size attracts momentum capital, and momentum capital leaves when the macro turns. March alone saw a fall of more than 10 percent, the sharpest monthly decline since 2013.

So the correction is better read as a repricing of positioning and rates, not a verdict on demand. Demand set records throughout.

The part most analysis gets wrong

Here is the distinction worth carrying away:

Eastern physical demand sets the floor. Western financial demand sets the price.

Chinese buying does not push gold up. It is not speculative, it is not leveraged, and it does not chase. What it does is absorb supply at lower prices, which changes the shape of the downside. A market with a large, patient, price insensitive buyer underneath it behaves differently in a selloff than one without.

That is a less exciting claim than "China is about to send gold to the moon," and it is a more useful one. It tells you that record Chinese imports are a reason to expect shallower corrections, not sharper rallies. The rally, when it comes, will be triggered by the Fed, the dollar, or a shock, and the floor will be the reason the market had somewhere to rally from.

What it means for prices from here

The honest answer is that the professional forecasts are extraordinarily far apart, and that spread is itself the most informative number in this article.

Year-end 2026 targets currently published by major banks:

From roughly USD 4,174 today, that is a set of forecasts ranging from a further fall to a near doubling. Several of those numbers were revised sharply upward during the year, which tells you how fast the view is moving.

Looking into 2027, Metals Focus projects an average near USD 5,330, and the bank consensus clusters in the USD 5,200 to 5,400 range.

What can be said with more confidence than any single target:

The structural bid is real and it is not going away. Central bank accumulation and Chinese household demand are responses to a reordering of reserve preferences and a shortage of domestic alternatives. Neither resolves in a quarter.

The cyclical headwind is the thing to watch, and it is visible. Rates and the dollar are observable in real time. You do not need a forecast to know which way they are pointing; you need to be watching them rather than the import data.

Record demand into a falling price is historically a constructive setup, not a bearish one. It means the physical market is clearing at these levels and inventory is moving east, where it tends not to come back.

Frequently asked questions

Westcourt Advisory is a cross-border advisory firm. This article is provided for general information and is current as at 14 October 2026. Price levels and forecasts cited are those published at the time of writing and will change. Nothing here constitutes investment, financial, legal or tax advice, or a recommendation to buy or sell any asset. Independent professional advice should be obtained before acting.