In an interview with Tim Treadgold for the Eureka Report, Mark Creasy makes a point about watching the stocks, not flow, of gold:

Scrap gold, and even mine-supply, aren’t really the big players in the (gold) market. There’s about 160,000 tonnes of gold in existence, and the world produces about 2000 tonnes of freshly-mined gold a year, and about 1000 tonnes is generated as scrap. The total amount of mined gold and scrap is less than 1% of the overall gold market. The mine supply isn’t all that important in the gold price. It’s all about sentiment. The people who will influence the value of that 160,000 tonnes are the biggest shareholders, and they are the central banks, which own about 30,000 tonnes.

Gold is a bit like a company which has a dominant shareholder. If everyone believes the dominant shareholder is selling the price drops like you wouldn’t believe. A major influence is how people see the biggest shareholders handling their gold.

…

The best way is look at gold is not on the peripheral, say the scrap market or even mine supply; it’s to ask what are the big shareholders doing. In the past we’ve seen big holders such as the Bank of England and the Swiss National Bank selling, and people think we’re out of this. When they see a big new buyer, people want in.

Now Mark’s point that the flows are peripheral is not to say that they do not have an influence on the price. In the long run, if we have passed a “peak gold” moment (a subject of an upcoming post) then flows reduce and stock levels out which, in the face of increasing demand, is bullish for price. Significant changes in flows can also tip the balance of the buying and selling volumes of the holders of the 160,000t stock.

Probably the biggest influence classical “flow” supply/demand analysis by WGC and GFMS has on price is via perception of its effect on price rather than any actual effect due to the physical volume of the metal. Consider that Mark’s point about stocks has one problem – how do you measure the position of gold “shareholders”? How much volume of gold will be bid and offered to the market at each price level? Even if you can poll all the holders, the so constructed supply/demand curves of the “stock” will change over time in response to events. This is Mark’s “sentiment”.

Now why “flow” supply/demand is so dominant is due to the fact that it is (relatively) nice and easy to measure. Sentiment is not. Therefore there is a bias towards the numbers, the measurable 3000t flow and away from (or complete disregard for) the unmeasurable 160,000t stock.

As Richard Maybury notes in his recent article:

… it is so important to see the economy not as a machine but as an ecology. Machines don’t feel, they don’t have fear, or joy, or optimism. But people, biological organisms, do have feelings. They do fear, and their fears can change instantaneously. The human ecology, especially these days, is driven very largely by emotions.

As an investor it is important to remember that while “flow” supply/demand numbers from WGC or GFMS do have an effect, it is not the entire story and one should be cautious of the resulting mechanistic analysis. If “fears change instantaneously” then the movement of the 160,000t stock will overwhelm the annual supply/demand balance.

There is a bit of circularity or feedback in that perception of annual flows impact sentiment of the holders of stock. Also note that instantaneous change can occur either way – if fear decreases then supply of the stocks will push the price down.

Big upward price moves will occur when the sentiment that gold stock holders have is what is commonly called “strong hands” coincides with “weak hands” of holders of dollars. Focusing on annual mine supply or a drop in Indian jewellery demand isn’t going to help you identify that move.