Gold and China’s weak premium trap

China’s Shanghai gold premium has shrunk to just $2.31 an ounce. Too many traders see that and declare Chinese demand dead. We think that is the wrong read. A smaller premium says the local market has caught up with demand, not that China has stopped absorbing gold, GoldSquawk data showed.

That bearish view is not theoretical. Pictet said Chinese imports and Swiss exports of gold, together with the Chinese gold premium, all point to weak physical demand from China. We think that is a lazy shortcut. The premium has cooled, yes, but the broader Chinese bid is still building through official buying, investment demand and rising metal held inside China.

A thinner premium is not the same as no demand

On Friday, Nov. 6, 2026, the Shanghai Gold Exchange benchmark was still above London spot by $2.31 an ounce, and the seven-day average premium was $6.78 an ounce, according to GoldSquawk data. That matters. China is still paying up versus London on average. The stress has eased; the market has not flipped into clear oversupply.

Yes, the latest reading was the lowest since Oct. 12 and the bottom of the recent $2.31-to-$12.5 range, according to GoldSquawk data. But that only tells us the local squeeze is softer than it was two weeks ago. It does not tell us the whole Chinese demand curve has fallen away.

This is where the premium-only argument goes wrong. A premium is a marginal signal. It is useful for showing whether nearby local metal is tight or loose versus London. It is not a full census of Chinese demand. It says little on its own about what official buyers are doing, what investors are doing, or how much metal has already been drawn into domestic hands.

The Chinese bid has changed shape, not vanished

The clearest rebuttal comes from the official sector. The State Administration of Foreign Exchange said China’s gold reserves rose to 78.30 million ounces at end-October, up 830,000 ounces on the month and marking a 24th straight monthly increase. Asharq Bloomberg put the October addition at 25.8 tonnes. That is not retreat. That is steady state buying.

Even before October was added, IMF International Reserves data showed China at 2,410 tonnes in September, up 103.3 tonnes so far in 2026 and sitting at the top of its one-year 2,304-tonne-to-2,410-tonne range after 23 straight monthly rises. If the world’s most watched official buyer keeps pushing holdings to the top of the annual range, a weak Shanghai premium is not a verdict on vanished appetite. It is a sign that one local indicator is being over-read.

The World Gold Council fills in the other half of the picture. It said Chinese jewellery demand stayed weak, but investment demand and official-sector buying remained strong. That is the split that matters. Some demand has weakened; other demand has strengthened. The mix has changed.

The WGC goes further. It expects investment demand to be the principal source of annual demand growth in 2026, at about 200 tonnes above 2025, and says China could remain the only major market to record annual growth, even if fourth-quarter demand does not match the year-to-date pace. That is not a market rolling over. It is a market rotating.

Rising stocks inside China are part of the story

The second lazy bear signal is warehouse metal. Gold in Shanghai Futures Exchange warehouses rose to 118.4 tonnes in the week to Friday, Nov. 6, 2026, up 1.5 tonnes on the week and the highest since Nov. 14, 2025, according to the Shanghai Futures Exchange weekly warehouse stock report. Bears will point to that and shout surplus.

We think that misses the obvious. More metal in Chinese warehouses means more metal has made it into the domestic system. Pair that with a still-positive local premium and another month of PBOC buying, and the cleaner reading is catch-up supply being absorbed, not demand disappearing.

The World Gold Council’s own China update pointed the same way. It said withdrawals from the Shanghai Gold Exchange were supported by seasonal factors while jewellery weakness weighed on demand, and it said a 23-tonne PBoC addition in the month it covered was the biggest since October 2023, taking third-quarter buying to 63 tonnes. Again, the shape of demand changed. It did not evaporate.

The broader market backdrop fits that argument. Across the 15 central banks tracked in IMF International Reserves data, net official gold buying this year was 216.4 tonnes, and seven of the 15 added gold in their latest reported month. China is not swimming against the official-sector tide. It is a leading part of it.

Even Friday’s macro tape did not produce the sort of collapse we would expect if a major physical constituency had walked away. US non-farm payrolls came in at 165,000 against a 65,000 forecast, average hourly earnings rose 0.4% month on month against a 0.3% forecast, and unemployment fell to 4.1% against a 4.2% forecast, according to FRED data in the calendar. Yet Comex gold still settled at roughly $4,075.80 an ounce on Friday, Nov. 6, 2026, up $2.70 on the day, after jumping $132.90 on Tuesday, Nov. 4, 2026, according to GoldSquawk data. That is not price action screaming that China has switched off.

What would really challenge this view

If we are wrong, the evidence will not be one more skinny premium print. It will be a sustained move into discount in Shanghai, a stop to PBOC buying, and softer signs of domestic absorption. The material in front of us shows the opposite: a positive premium, higher Shanghai warehouse stocks and a 24th straight month of official buying.

The next hard test comes on Tuesday, Nov. 10, 2026, with US CPI, core CPI and monthly CPI due at 13:30 UTC. A hot print could hit futures and squeeze the Shanghai premium further. But if that macro pressure lands and China still keeps the premium positive while official reserves keep climbing, the weak-premium bear case will look exactly what it is: a thin signal stretched far past its worth.