Gold was up today after a long slide last week. The precious metal has decreased significantly in price since Goldman Sachs recommended investors short the asset. Today Goldman is out with a new report on bubbles. While the report focuses on bonds, Goldman devotes a short section to gold. Goldman concludes that gold (at least the ETF) could possibly be classified as a bubble. Below is the excerpt on the commodity followed by the full document in scribd:
The recent pronounced decline in gold prices clearly brings into question whether the gold market is a bubble in the midst of bursting. Even before this recent sharp pull back in gold prices, market fundamentals had already raised concerns around whether gold was a bubble. At the center of this concern was the physical backed gold ETF. At its peak at the end of 2012 its holdings were 84.6 million oz. This made it one of the world’s largest gold reserves behind only the German and US central banks and the IMF. Since then it has seen a sharp decline of 9.4 million toz., which alone is 10% of annual gold production. But with this drop only Italy and France have barely edged it out.
Defining a bubble
The textbook definition of a bubble is asset price inflation generated by excess demand, which is not fundamentally supported. The ETF could clearly be described as this type of excess demand. The key question is whether this excess demand leads to a buildup in long positions that can quickly unwind and crush prices in the future, i.e., popping the “bubble.” Because the gold ETF is backed by physical gold, you could also similarly ask whether the ETF holdings have literally generated an inventory build of gold that can come back into the market and push prices sharply lower. We believe that the ETF is an inventory build that has this bubble property, leaving the only question of how “sticky” is this inventory?
Surplus versus hoarding
An inventory build that is associated with declining prices is usually labeled a “surplus.” Such an inventory build typically ends when prices fall low enough to create a fundamental correction, which in turn draws the inventory down, causing a price rise in the future. In contrast, an inventory build that is associated with rising prices is usually labeled as “hoarding.” Hoarding usually ends when the hoarder either proves to have been prudent, in which case the realization of the event that motivated the hoarding in the first place causes the inventory to decline, or it becomes clear that the event that motivated the hoarding will not occur, in which case the inventory re-enters the market and crushes prices.
During the US housing boom people hoarded property under the belief that demand and prices would continue to rise. When that belief was proved wrong, the inventory re-entered the market and prices collapsed. Interestingly, we never use the term “bubble” to describe this dynamic in food and energy when “precautionary” inventories are built in anticipation of events such as bad weather that are expected to create demand for the inventory and cause
prices to rise. Instead, we call it “prudent” despite the fact that it is very difficult to predict these events or even assess the probability of them occurring. However, when that bad weather event doesn’t happen – just like a bubble – the precautionary inventories re-enter the market and crush prices. The only reason we call the precautionary inventory build in food and energy prudent is that it determines our mortality unlike houses, gold or bonds.
Fundamental drivers, but hard to pin down
In gold, the weather event is a systemic financial or monetary crisis and the ETF can be interpreted as the precautionary inventory build to protect from this event. In recent years we have linked changes in ETF demand to movements in sovereign credit default swaps, real interest rates and exchange rates – all measures of debasement and the potential risk for a substantial financial crisis. Unfortunately, these correlations are not consistent over time, which makes it extremely difficult to connect the level of the ETF to a measurable probability of sovereign default, debasement or other financial crises. And because this relationship changes over time, it is extremely hard to predict despite our view that it is entirely rational behavior.
How sticky is ETF demand?
So just how sticky is the ETF demand, driven by a diverse group of small and large retail, institutional, and sovereign investors? As the economic climate improved during 1Q13, ETF holdings declined in the largest and longest stretch since inception. Given the sharp drop in prices, we estimate that 12.7% of the existing holdings, or 9.55 million oz., have a cost basis above current levels, pointing to potential further reduction in under-water holdings. And while using real interest rates as a guide to how ETF holdings could evolve going forward points to an additional 7.6 million oz. decline in ETF holdings by year-end 2014, the recent pace of decline has been much stronger, pointing to downside risk to our $1270/toz. year end 2014 target. As we know that these relationships were unstable on the way up, there is no way to be certain it doesn’t all rush to the exit at once. The history of commodity markets tells us that when a precautionary inventory build is deemed unnecessary it nearly always proves not to be sticky.