I spent the past week in Dubai – City of Gold – a hub for both bullion and jewellery trade. Gold Jewellery here is sold by weight and purity, with most of the jewellery being either 22 karat or 24 karat gold (24 karat gold is pure gold; 22 karat is 22/24ths pure, or 91.67%). In contrast, much of the jewellery sold in the United States is either 10 karat (41.67% gold) or 14 karat (58.3% gold). 22 and 24 karat gold jewellery is soft and impractical for everyday wear, which is why most of the jewellery in the West contains considerably less gold.
Why gold jewellery demand doesn’t matter
Jewellery containing a high percentage of gold is not purchased for everyday wear but for its value as a store of wealth. Because the manufacturing cost component of the final price for jewellery with a very high gold content (22 and 24 karat) is much smaller than the manufacturing cost component for jewellery with a low gold content, high purity jewellery can be resold to jewellers for its gold content with very little financial loss. Such sales make up the bulk of the gold recycling supply.
When you are in a place such as Dubai, where so much trade in physical gold takes place, it is easy to understand why people think jewellery production is important to the gold price. However, the price of gold is almost completely insensitive to jewellery demand, scrap sales, mine production, producer hedging or central bank sales.
Total jewellery consumption during 2005 was 2,736 tonnes of gold. Industrial and dental demand added another 419 tonnes for a total of 3,155 tonnes of primary gold demand. Mine production during 2005 was about 2,515 tonnes. Some people believe that this 640 tonne primary shortfall of gold mine production relative to fabrication demand is relevant to the gold price. They also seem to think that net investment demand, official sector sales and purchases, etc. are important. Jewellery demand is important to jewellers but it makes no difference to the gold price.
Gold is money
The bulk of international gold trading occurs through the facilities of the London Bullion Market Association (LBMA) and from their website we learn that the average DAILY volume of gold trading was 513 tonnes during 2005. Now let me explain what that number means. The LBMA only reports the net amount of gold transferred between its clearing members at the end of the day. So if two members trade 100 tonnes five times during the day (total trading volume is therefore 500 tonnes), and at the end of the day one member owes the other member 200 tonnes of gold (i.e. net position at the end of the day) then only 200 tonnes is reported by the LBMA. The LBMA trading volume reflects the net amount of gold transferred between clearing member accounts at the end of the trading day. It is estimated that the actual volume of gold trading is possibly as much as 4 to 6 times larger than the net positions reported by the LBMA. Remember, these are not derivative trades; these are the settlement numbers for actual gold trades.
Now, if the volume of gold trading through the facilities of the LBMA is in the order of 500 tonnes per day, then I am afraid that a 640 tonne annual shortfall between mine production and fabrication demand is totally irrelevant. In fact, the entire 2,736 tonnes of annual jewellery demand are irrelevant: they represent less than 6 days of trading through the LBMA. It makes no difference to the gold price whether India had a good harvest or a poor harvest, even though, again, it is probably important to the jewellers.
Gold is money; it is not a commodity. Its value relative to fiat currencies is purely a function of the relative inflation rates of gold and other currencies. The gold price may fluctuate based on investor psychology, just like the price of any financial asset, but it will always revert back to its fundamental value, which is determined by relative inflation rates.
Just like the inflation rate of money is the increase in money supply, the inflation rate of gold is the increase in the above-ground stock of gold, i.e. mine production as a percentage of all the gold that has ever been mined. If we know what the inflation rate of a particular currency is we can calculate its relative value to gold, assuming we have a starting point at which we knew what the fair value of the currency was. This can be done for the US dollar. In 1933 the gold price was defined by the fact that a $20 gold coin contained 0.9675 ounces of gold. From this we know that the gold price in US dollars during 1933 was $20.67. If we now take in consideration the inflation rate of the US dollar (as defined by M3 and the Consumer Price Index for those years that M3 did not exist), and the inflation rate of gold (annual mine production) we can calculate that the gold price in US dollars should be around $850 an ounce.
How much is an ounce gold really worth?
However, gold is not unique to the United States; it is an international form of money and so we also have to take into consideration the US dollar exchange rate, since anything we price on international markets in US dollars will fluctuate along with changes in the US dollar exchange rate. We know from the US trade deficit that the US dollar is over-priced on foreign exchange markets and the US is putting considerable pressure on China to help devalue the dollar. The only reason gold is not $850 an ounce (or thereabout) today is that the US dollar is over-priced. As the US dollar exchange rate falls the gold price in US dollars will rise.
Like all markets there is a real possibility that the gold price will overshoot its fair value and the gold price could therefore exceed $850 an ounce. But like all markets the gold price is bound to return to its fair market value in the event that it does overshoot the mark. Also, keep in mind that the value of gold in US dollars will increase over time in proportion to the inflation rate of the US dollar. Unfortunately, the Federal Reserve of the United States has stopped publishing M3 data and so we have lost a great tool for gauging the dollar’s inflation rate. Historically the inflation rate of the dollar has almost always outpaced the inflation rate of gold, so the longer it takes before the dollar exchange rate falls, the higher the fair value of gold in US dollars becomes.
None of what I wrote here is new. I have been writing about the fallacy of viewing gold as a commodity since 1998 (the LBMA announced the volume of gold trading for the first time in late 1997). Below are links to older articles that expand on this week’s ideas if you would like further reading.
https://www.paulvaneeden.com/Library/200310%20Commodity.php
https://www.paulvaneeden.com/Library/200304%20Gold.php
First published on Kitco.com
By Paul van Eeden
Paul van Eeden works primarily to find investments for his own portfolio and shares his investment ideas with subscribers to his weekly investment publication.