Serious Banks Are Forecasting Gold at $4,500 and at $6,000 for the Same December

J.P. Morgan has looked for around 755 tonnes of central bank purchases this year, with other houses clustering in a 750 to 850 tonne range. The World Gold Council's survey work has found that around 95 per cent of central banks expect global official gold reserves to grow over the next twelve months, and roughly 43 per cent intend to add to their own.

That matters more than any fund flow, for a simple structural reason: central banks are close to price insensitive. They are not trading. They are reallocating reserves over years for reasons that have nothing to do with this quarter's chart, and once bought, that gold does not come back to market on a 10 per cent rally.

It puts a slow, heavy bid under the price. It does not stop the price falling 28 per cent, as January proved.

What is actually moving gold right now?

A tug of war, and both ends of the rope got shorter this month.

Pulling up: geopolitical risk, with Brent above $91 after the US and Iran failed to extend their interim arrangement, and continued official sector buying.

Pulling down: yields. The US thirty year Treasury is around 5.32 per cent and the ten year near 4.74 per cent, the highest in twenty months. A metal that pays you nothing competes badly with a government bond that pays you over five per cent, and that arithmetic has not been this unfavourable to gold in two decades. Our piece on why long bond yields are at multi-decade highs covers the other side of that trade.

Gold is up 9.5 per cent this month anyway. When an asset rises against its own headwind, the honest conclusion is that somebody is buying it for a reason the model does not contain.

How should you read a gold price target?

Four questions, and they will disqualify most of what you see.

1. What is the horizon, and does the number come with a range? A single figure for December, with no distribution around it, is marketing.
2. Which buyer does it assume? A $6,000 forecast is a forecast about official sector demand and fiscal anxiety. A $4,500 forecast is usually a forecast about real interest rates. They are not the same argument and should not be compared as though they were.
3. Has it been revised, and how quietly? Targets get moved. A house that has revised three times this year is telling you something about its conviction that the current number does not.
4. What would make it wrong? A forecast that cannot be falsified is a slogan.

Gold at $4,390 is neither cheap nor obviously expensive. It is an asset in the middle of a genuine argument between people who are paid to know, and the correct response to that is smaller positions and longer horizons rather than a stronger opinion. Our risk map for the second half sets out what would have to happen for each of these numbers to be the right one.