Jul 172015

In March 2015 the CME launched a gold kilobar futures contract. As with all futures contracts, vaulters apply to be a warehouse and as part of that they have to report registered and eligible stocks in their vaults. Currently there are three vaults reporting figures:

  • Brinks – 820,204 ounces
  • Malca-Amit – 36,909 ounces
  • Loomis International (ie Via Mat) – 17,779 ounces

The figures above are pretty representative of the average balances held by these three since the contract started trading. It is interesting that the Hong Kong warehouses have never reported any registered stock – it is all eligible. Compared to the CME’s US warehouses however, the 874,893 ounces of gold held within the Hong Kong warehouses is only 10% of the US stock of 8,751,688.

I hadn’t given the Hong Kong contract or its warehouses much thought until Ronan Manly, who writes for BullionStar, drew my attention to this submission by the CME to the CFTC “self-certifying the listing of a Gold Kilo Futures contract”.

As part of that submission the CME provides an analysis of deliverable supply so as to determine a conservative spot month position limit. To do that analysis “the Depositories that are intended to be approved by the Exchange … provided historical inventory levels of gold kilo bars stored in their respective vaults that meet the specifications of the Gold Kilo futures contract.” Those two depositories were Brinks and Malca-Amit.

The chart below shows the result of combining the data from the CME’s submission (monthly averages) with reported warehouse stocks for the HK contract supplied by Nick Laird at Sharelynx.com – note that the submission data only goes up to November 2014 and the contract starts trading in March, so we have a gap in our data.


You can see that someone(s) was accumulating kilobars all through 2013 and 2014, with the major surges being between Dec 2012 to Feb 2013 and Nov 2013 to Jan 2014. The interesting thing about that gap in our data is the massive drop of about 110 tonnes in the space of three and a half months. Where did it go?

The first theory may be that it went to the SGE to meet Chinese demand. However looking at exports from Hong Kong to China during that period shows no significant changes in volumes. There is also no indication that Chinese demand surged during that period, with SGE premiums (an indicator of demand) remaining subdued and SGE withdrawals at a high but relatively consistent rate.

The other possibility is that the owner(s) of that 110 tonnes realised that come March 2015 their previously unreported (at least until the CME submission) hoard of gold would suddenly become public, as CME futures warehouses, according to the CME rulebook, are (my emphasis) “required to report inventory to the Exchange .. Eligible metal shall mean all such metal that is acceptable for delivery against the applicable metal futures contract for which a warrant has not been issued.”

The purpose of that rule is to give the market visibility into potential stocks that may be traded on the contract, and also to avoid games where someone could take metal off warrant (no longer be registered) to give the impression there is little stock available to the market in the hope of giving the impression of a shortage.

The only way the owner(s) of that 110 tonnes could avoid be reported would be to move the metal into a non-CME warehouse, it being my assumption that the CME does not give exemptions to its warehouses on reporting the stocks of individual owners.

The chart below shows the combined Hong Kong and US warehouse data from Sharelynx.com. Note the data gap marked (1), which shows how significant the 110 tonne reduction was in terms of overall CME warehouse stocks. It is interesting to note that the area marked (2) begs the interpretation that metal was moved from the US to Hong Kong, although my cursory analysis of US-Hong Kong imports and US warehouse withdrawals does not confirm this.


I noted previously that the rapid “decline in registered gold stocks and delivery rates occurred soon after gold’s dramatic crash through $1550 and into the 1300s” and it is interesting to note that the rebuild of overall stock (in Hong Kong rather than the US) appears to have begun after gold bottomed for the second time at $1200 in December 2013.

One final note on the CME submission. As mentioned, the CME used Brink’s and Malca-Amit’s data to determine a “conservative” position limit: “As the basis for assessment of deliverable supply, the average monthly combined gold kilo bar inventory … is 89,408 kilo bars. Staff proposes a conservative spot month position limit of 6,000 contracts which is 6.71% of deliverable supply.”

Given the massive drop in Malca-Amit’s warehouse stock, I think it could be argued that a review of those position limits is warranted. The average Hong Kong warehouse balances since the start of the contract’s trading has been 27,730 kilobars (with a minimum of 17,296 bars and a maximum of 40,587 bars), which at 6.71% of deliverable supply would be 1,860 contracts, which we can round up to  2,000 contracts – compared to the current 6,000 contracts.