December 8, 2003 – Since the spring of 2002 the World Gold Council (WGC) has been developing an exchange-traded fund (ETF) for gold. The objective is to eliminate the hassle and inconvenience of handling physical metal by providing the ownership of gold through a listed security that can be conveniently purchased and sold through stockbrokers.
I have been following the WGC’s efforts closely because I wanted to see if their ETF would have a comparable level of governance to what my colleagues and I have achieved in GoldMoney. A product launched by the WGC could have some competitive impact. Additionally, GoldMoney has been exploring the possibility of creating its own ETF using goldgrams as the underlying asset.
In May 2003 the WGC submitted to the Securities & Exchange Commission (SEC) the S-1 registration statement required to obtain approval for a New York Stock Exchange listing. Subsequently, two amended versions of the S-1 have been submitted to the SEC, the most recent being last month. Why hasn’t the filing been accepted by the SEC?
The SEC doesn’t disclose the reasons for its actions, and the WGC isn’t talking. But a lot can be gleaned from reading the WGC’s most recent SEC filing as well as the prospectus for a fund that the WGC plans to launch Tuesday, December 9th in London. The SEC may be on to something meaningful and important.
Both of these WGC funds have very loose custodial controls. This custodial structure opens the possibility that the fund is not completed supported by physical metal. In other words, the fund may have fractional reserves, which would of course mean that shareholders to some extent could be buying paper promises rather than physical metal.
This possibility was first raised in an article by Tim Wood of MineWeb.com that was posted on May 15, 2003, soon after the S-1 was first made public. Tim noted the shortcomings of the proposed WGC fund by comparing it to GoldMoney. See: www.mips1.net
Even though he doesn’t mention GoldMoney by name, Tim comments appear to be directed toward us and to one of our strengths – exceptional governance procedures that mitigate risk and provide customers with assurances of integrity. He wrote: “A notable feature of digital gold operations is their fetish over the gold entrusted to them in terms of ensuring its purity and location.” He then notes that those same controls are lacking in the WGC’s ETF: “Whilst a problem is unimaginable given the players involved, it is nevertheless odd that Equity Gold has been unable to secure iron clad (forgive the mixed metaphors) insurance, purity, creditor and custodianship guarantees. It might be nitpicking, but it does illustrate the issue of counterparty risk which does not vanish just because the underlying product is gold.”
There is a problem here that needs fixing. The custodial control of the WGC funds needs to be strengthened. These funds need to provide assurances about the physical metal segregated in allocated storage. As these funds are now structured, these assurances are lacking because the custodian has no control over the sub-custodians or the sub-sub-custodians.
To illustrate my concerns, I include below quotes from the prospectus of the UK fund that is about to begin trading, Gold Bullion Securities Limited, which can be downloaded from: http://www.goldbullion.com
p 11- “The gold will be held in custody by the Custodian, sub-custodians or their delegates…”
p 22- “All Trust Gold will be held by the Custodian at its London vault premises or in the vaults of any sub-custodian or by a delegate of a sub-custodian.”
Interestingly, the prospectus doesn’t list the location of the vaults of the sub-custodians or their delegates. What is the political risk in those countries where those vaults are located? How secure are the vaults of the sub-custodians or their delegates? Even more importantly, who are these delegates?
Further, the Custodian is not held accountable for the sub-custodians or their delegates.
p 38 – “The Custodian Agreement requires the Custodian to use reasonable care in the selection of those sub-custodians and provided that is (sic) shall not be liable for any act or omission, or for the solvency, of any sub-custodian it appoints unless the appointment of that sub-custodian was made by it negligently or in bad faith.”
p 38 – “The Custodian is under no duty or obligation to make or take, or require any sub-custodian it appoints to make or take, any special arrangements or precautions beyond those required by any applicable rules of the LBMA, the Bank of England or any other applicable regulatory authority.”
But the prospectus makes clear that there is no regulatory supervision over custodial services in London.
p 28 – “Custodial services offered by the Custodian and any sub-custodian are presently not a regulated investment activity subject to the supervision and rules of the Financial Services Authority, the United Kingdoms financial services regulator (or the Bank of England).” And as noted in the MineWeb article, the fund does not have any requirement that the gold be insured.
p 27 – “There is a risk that the Trust Gold could be lost, stolen or damaged…If the Custodian fails to take out suitable insurance then Security Holders may have to rely on having a claim against the Custodian. The Custodians liability is limited in various ways.”
So there is a lot of risk here. Is the gold really in the vault? Or is some sub-custodian or delegate just accepting dollars from people buying shares in exchange for paper? Given this existing custodial arrangement, there is no way of knowing for sure. Consequently, it is reasonable to conclude that this deficient custodial control may explain why the SEC has not accepted the WGC’s proposed US fund.
The US fund – like its UK counterpart – has the same loose custodial structure because not only the Custodian, but also the subcustodians, can handle the gold owned by the US fund. And the fund’s Trustee has no control over the subcustodians. The following quote is taken from the WGC’s proposed US prospectus:
“The Trustee does not undertake to monitor the performance of any subcustodian. Furthermore, the Trustee may have no right to visit the premises of any subcustodian for the purposes of examining the Trust’s gold or any records maintained by the subcustodian, and any subcustodian may not be obligated to cooperate in any review the Trustee may wish to conduct of the facilities, procedures, records or creditworthiness of such subcustodian. In addition, the ability of the Trustee to monitor the performance of the Custodian may be limited because under the Allocated Bullion Account Agreement and the Unallocated Bullion Account Agreement (together, the Custody Agreements) the Trustee has only limited rights to visit the premises of the Custodian for the purpose of examining the Trust’s gold and certain related records maintained by the Custodian.”
Why should a gold ETF be different from any other exchange-traded fund already registered with the SEC? These other ETFs provide for complete control over the custodian as well as all sub-custodians. It is only through such control that shareholders can be certain that a fund owns physical gold that is safely and securely stored in a vault.
I assume the WGC has structured their funds this way in recognition of how the bullion market operates. The processes for trading in the bullion market are based upon one basic principle – moving paper around is easier and less expensive than moving metal. So these funds are structured to accommodate the way bullion banks now operate, therefore providing the WGC funds access to the liquidity these banks offer as well as avoiding the imposition of the fees that would result if the banks were continually moving metal between themselves instead of paper.
Basically, the London bullion banks operate using 19th century procedures. Their procedures never entered the 20th century, let alone the 21st. The core principle of these procedures is ‘my word is my bond’, which may work in some circumstances, but probably does not work where fund managers have an obligation to carry out due diligence for their clients and/or where investors want assurances that when they are buying something reported to be gold, that the gold in fact is actually there.
In short, these WGC funds may in practice only be fractional reserves, and not 100% gold. The prospectus allows that possibility, and discloses the risks relating to it. Further, the UK and US prospectus each ring-fence the liability of the custodian and the trustee, relieving them of any responsibility if it turns out that there is ‘fiddling’ by any subcustodian. So it seems understandable that the SEC would be reluctant to accept for filing the WGC prospectus.
Lastly, in the past I have had no concerns about the WGC’s Australian gold fund, which has been operating for a few months. However, in view of these new discoveries about the uncertainty surrounding the custodial arrangements of the other WGC funds, I have now also looked again at the prospectus of the Australian fund.
Clearly, the Australian authorities seem to allow a lot less disclosure than the regulatory authorities in the US, or even the UK for that matter. So it is hard to draw any clear conclusions about the safety of assets in custody, which in itself is a concern. But the prospectus for the Australian fund does allows the Custodian to appoint subcustodians, which presumably creates the same risk that exists in the WGC’s proposed US and UK funds. Therefore, based on my reading of the Australian prospectus in view of the risks disclosed in the UK and US filings, I cannot recommend the WGC’s Australian fund.
The risks of the WGC’s funds appear too great. Until more questions are answered and/or the fund’s structure is changed to eliminate its loose custodial controls, I do not recommend that these funds be purchased.