Where Is the Copper Supply Surplus Gone?

Only less than two months ago, the International Copper Study Group forecasted at its semi-annual meeting in Lisbon that after four years of supply deficit, the copper market would swing into a surplus. While the ICSG estimates that the apparent refined copper consumption exceeded supply in 2013 by 282 thousand tons. For this year, the ICSG forecasts a supply surplus of 405 thousand tons. This swing in the supply/demand-balance by 687 thousand tons is due to an increase in mine production, which leads also to a rise of refined copper output. For this year, global mine production is expected to increase by 4.7% to 18,904 thousand tons. Refined copper production is forecast to rise faster than mine production by 6.5% to 22,362 thousand tons. The global copper consumption is predicted to increase by only 3%, which is less than global GDP growth, to 21,957 thousand tons.

A swing in the supply/demand balance of the magnitude predicted by the ICSG should be reflected in the development of copper stocks. According to the recent figures published on May 22, 2014, by the ICSG in the press release for the Monthly Copper Bulletin, global copper stocks increased by a total of 102 thousand tons in the first two months of 2014. Inventories held in the warehouses of the three major exchanges (LME, Shanghai and CME) dropped by 25.2 thousand tons in the same period, which is in a clear contrast to the estimate of the ICSG for the first two month. Since March, the fall of inventories accelerated. Despite some temporary inventory builds at some exchanges, copper stocks are currently down 228 thousand tons YTD. This plunge of copper inventories at exchange warehouses since the beginning of March is in such an obvious contrast to what one would expect given the forecast of a 405 thousand tons supply surplus.

Certainly, copper inventories are not only held in exchange warehouses. Producers and consumers of copper also hold inventories, but one would expect that these holdings are more determined by economic activity and opportunity costs. If we take this into account, it appears rather unlikely that producers and consumers have increased their inventory holding by the amount to explain the difference between the change in the supply/demand-balance and exchange warehouse inventories.

Furthermore, if producers hold more inventories, then one would expect that they also want to hedge the inventories, especially when the majority of forecasters predict a change in the supply/demand balance. This should lead to a considerable increase of open interest and falling copper prices. However, when copper reached its low of the year so far in March, the LME copper futures open interest was up only 7 thousand contracts compared to the end of 2013. In addition, copper at the LME remained in a slight backwardation in mid-March. But copper prices were in contango at the Shanghai Futures Exchange. LME open interest in copper futures rose far stronger since mid-March to more than 340 thousand contracts at the start of this past week. This rise in open interest was accompanied by a recovery of the copper price. This argues that not producers are hedging against further falling prices due to a supply glut but consumers hedge future demand. However, this behavior would be only rational if consumers doubt that forecasts of a supply surplus depressing prices are correct.

The other explanation is that the copper supply surplus is held in bonded warehouses in China and had been used for financing deals. In April, we analyzed the development of the copper price in Shanghai and the inventories held at SHFE warehouses. In 2012 and 2013, a structural break took place and inventories at SHFE warehouses rose stronger as a model based on the development between 2009 and 2011 would have predicted. However, this development had not a major impact on the price of copper. Other factors explained the price development far better. It could not be ruled out, that the higher SHFE warehouse inventories were the result of stocks moved out of bonded warehouses.

Thus, doubts remain that the excess production of refined copper was absorbed by stocks held in bonded warehouses in China for serving as collateral in financing deals. Furthermore, it could not explain the decline of stock held in exchange warehouse inventories

While total exchange warehouse inventories decayed since the start of this year, stock held in LME warehouses declined by 195 thousand tons and remained almost unchanged in CME warehouses. At SHFE warehouses, inventories rose until the plunge of copper prices, and then also sunk from 213.3 to 92.7 thousand tons, which translates to a fall of 33.2 thousand tons since the end of 2013. It had been reported at the end of April that China’s State Reserve Bureau bought copper. However, according to the sources of the reports, the SRB bought from bonded warehouses. But this does not explain the fall of copper stocks at SHFE warehouses, unless the SRB also bought silently since mid-March copper in huge amounts held in bonded and exchange warehouses.

We have data for cancelled warrants only for LME warehouses. At the LME, the free available copper stocks plunged to a mere 92,650 tons at the end of this week. This is the lowest level since early April 2008. Copper is in a backwardation at the LME and the SHFE, which indicates that copper is tight and not abundantly available as estimates about a supply surplus would suggest. Our quantitative model for global total refined copper supply and consumption indicates that consumption would further exceed supply in this year. Structural changes can have a significant impact, which quantitative models do not incorporate immediately. Thus, we would not rely solely on quantitative forecasting models. However, the development of copper inventories in exchange warehouses, the backwardation and the results of the quantitative model increase our skepticism that the copper market will swing to a supply surplus in the magnitude forecasted by the ICSG.