Imagine the following: you read in a newspaper that a group of investors has sold U.S. dollars to the tune of $820 million over the past two months for other currencies. This incidentally represents approximately 0.082% of the broad dollar money supply TMS-2 (which amounts to roughly $9.3 trillion at present). It means they would have been selling roughly $20 million per trading day. You then learn that $4 trillion of U.S. dollars are traded in global currency markets every single trading day. Would you believe that their selling has influenced the exchange value of the dollar beyond a rounding error?
And yet, we are supposed to believe that the selling of an equivalent amount of gold from the gold holdings of exchange traded funds over the past two months (they have sold 140 tons, or 0.082% of the total global gold supply) has greatly influenced the gold price.
According to a recent press report:
“Exchange traded funds have transformed the gold market. Since the first fund was launched nearly a decade ago, the products have become so successful in offering a simple way for investors to buy physical gold that they have acquired the nickname “the people’s central bank.”
But what happens when the people’s central bank decides to sell?
That is the question now haunting the bullion market. Since the start of January, gold ETFs have dumped 140 tonnes of gold. February saw the largest monthly outflow of gold from ETFs on record.
The sell-off is partly a reflection of broader negative sentiment towards gold, as investors become more confident in the global economy and put their money into riskier assets such as equities. Prices have slid 12 per cent since October to less than $1,580 (U.S.) an ounce, and are down 18 per cent from their record nominal high in 2011.
But the shift to selling by ETF investors is a concern in its own right for the gold market.
“The acceleration in gold ETF outflows is worrying,” says Joni Teves, precious metals strategist at UBS. Suki Cooper, analyst at Barclays in New York, describes a continuation of the current selling trend as “the key downside risk for prices.”
The reason is that ETFs have become a major presence in the physical gold markets.
Since the first gold ETF was launched in Australia in 2003, the products have become enormously successful. The funds offer investors a relatively low cost and easily tradable way to access physical gold by holding gold in a trust, which then issues shares that can be traded on an exchange.
With some 2,491 tonnes of gold held around the world, collective ETF holdings outnumber all but two central banks: the U.S. and Germany. ETFs hold enough gold single-handedly to supply the jewellery market in India – the world’s largest consumer of the metal – for more than four years.”
Sounds real scary, right? However, the only sentence above that is in any way meaningful is this one: “The sell-off is partly a reflection of broader negative sentiment towards gold.”
How the Gold Price Forms
That is all it is: a reflection of sentiment. The tonnages traded and the fact that ETFs could theoretically supply “Indian jewelry demand for four years” are more or less irrelevant. What India consumes in jewelry in four years is the amount of gold traded in London alone on three or four good days. Since all the gold that exists is always held by someone, the total demand for gold is at any moment equal to its total supply: it amounts to 170,000 tons. 140 tons sold by ETFs in two months are like a drop in the ocean.
As Robert Blumen has explained in great detail in an article we have published here a while ago (“What Determines the Price of Gold?”), the decisive factor for the gold price is neither mine supply, nor scrap supply, nor jewelry demand, and by extension, also not the supply provided and the demand exercised by ETFs. Instead it is the reservation demand of the current holders of gold that is the most important factor determining the price of gold.
So how can we, given these facts, make educated guesses as to where the gold price will head, or whether gold is presently over- or undervalued? To this end one must use indirect means, namely one must consider the fundamental drivers that are apt to induce gold holders to either lower their selling price or hold out for a higher one. We have discussed these in previous articles, so won't go into these details again here. See “Gold Still Misunderstood”, “Precious Metals, an Update”, “Some Thoughts on Gold Part One and Part Two” for a discussion of these drivers and how to properly analyze gold.
Having said all that, we want to briefly return to the use of ETF inflows and outflows as a sentiment indicator. It is true that a persistent continuation of such outflows would be a worrisome sign, as it would indicate waning investor interest in gold. Although the amounts concerned are too small to matter for the gold market, the ETFs are a reasonably good microcosm of the sentiment toward gold. However, experience has shown that large outflows in the very short term are often a contrary indicator. The ETFs tend to be the last entities where selling materializes, and selling in size is often seen near the end of short term downtrends. Should the holdings of gold ETFs stabilize here and inflows resume, we would have confirmation that investor sentiment toward gold is once again improving. It is solely in this sense it is useful to monitor these flows.