An article in February on BullionStar’s website titled “A Chink of Light into London’s Gold Vaults?” discussed an upcoming development in the London Gold Market, namely that both the Bank of England (BoE) and the commercial gold vault providers in London planned to begin publishing regular data on the quantity of physical gold actually stored in their gold vaults.
Critically, this physical gold stored at both the Bank of England vaults and the commercial London vaults underpins the gargantuan trading volumes of the London Gold Market and the same market’s ‘liquidity’. Therefore, a new vault holdings dataset would be a very useful reference point for relating to London’s ‘gold’ trading volumes as well as relating to data such as the level and direction of the gold price, the volume of gold held in gold-backed Exchange Traded Funds (ETFs), UK gold import and export statistics, and Swiss and Hong Kong gold imports and exports.
The impending publication of this new gold vault data was initially signalled by two sources. Firstly, in early February, the Financial Times (FT) wrote a story claiming that the London Bullion Market Association (LBMA) planned to begin publishing 3 month lagged physical gold storage data for the entire London gold vaulting network, that would, according to the FT:
“show gold bars held by the BoE, the gold clearing banks, and those [vaults] operated by the security companies such as Brink’s, which are also members of the LBMA.”
The “gold clearing banks” are the bullion bank members of London Precious Metals Clearing Limited (LPMCL), namely, HSBC, JP Morgan, ICBC Standard Bank, Bank of Nova Scotia – Scotia Mocatta, and UBS. HSBC and JP Morgan operate precious metals vaults in London. See profile of JP Morgan’s London vault and a discussion of the HSBC vault . ICBC Standard Bank also maintains a vault in London which is operated on its behalf by Brinks.
The second publication to address the new gold vault data was the World Gold Council. On 16 February, addressing just the Bank of England vaults, the World Gold Council wrote in its Gold Investor publication that:
“The Bank of England is, for the first time, publishing monthly data revealing the amount of gold it holds on behalf of other central banks.”
“The data reveals the total weight of gold held within the Bank of England’s vaults and includes five years of historical data.”
While I had been told by a media source that the London vault data would be released in the first quarter of 2017, at the time of writing, there is still no sign of any LBMA vault holdings data covering the commercial vault operators in London. However, the Bank of England has now gone ahead and independently released its own numbers covering gold held in the Bank of England gold vaults. These gold vaults, of which there are between 8 – 10 (the number fluctuates), are located on the 2 basement levels of the Bank of England headquarters in the City of London.
In an updated web page on the Bank of England’s website simply titled ‘Gold’, the Bank of England has now added a section titled ‘Bank of England Gold Holdings’ and has uploaded an Excel spreadsheet which contains end-of-month gold holdings data covering every month for a 6-year period up to the end of December 2016, i.e. every month from January 2011 to December 2016 i.e. 72 months.
According to the Bank of England, the data in the spreadsheet shows:
“the weight of gold held in custody on the last business day of each month. We publish the data with a minimum three-month lag.
Values are given in thousands of fine troy ounces. Fine troy ounces denote only the pure gold content of a bar.
We only accept bars which comply with London Bullion Market Association (LBMA) London Good Delivery (LGD) standards. LGD bars must meet a certain minimum fineness and weight. A typical gold bar weighs around 400 oz.
Historic data on our gold custody holdings can be found in our Annual Report.”
Prior to this spreadsheet becoming available, the Bank of England only ever divulged gold vault quantity data once a year within its Annual Report, for year-end reporting date end of February.
You will appreciate that the new spreadsheet, having data for every month of the year, and for 72 months of data retrospectively, conveys a lot more information than having just one snapshot number per year in an annual report. Therefore, the Bank of England has gone some way towards improving transparency in this area.
Before looking at the new data and what it reveals, it’s important to know what this data relates to. The Bank of England provides gold custody (storage) services to both central banks and a number of large commercial banks. Large commercial banks which trade gold are commonly known as bullion banks, and are mostly the high-profile and well-known investment banks.
On its gold web page, the Bank highlights this fact – that it provides gold custody service to both central banks and commercial banks:
“We provide safe custody for the United Kingdom’s gold reserves, and for other central banks. This supports financial stability by providing central banks with access to the liquidity of the London gold market.
We also provide gold accounts to certain commercial firms that facilitate access for central banks to the London gold market.”
In the London Gold Market, the word “liquidity” is a euphemism for gold loans, gold swaps, and gold trading including gold sales. This reference to central banks accessing the London Gold Market as being in some way supportive of ‘financial stability’ is also an eye-opener, since reading between the lines, the Bank of England is conceding that by accessing the London Gold Market’s “liquidity” via bullion banks, these central bank clients are either contributing to direct stabilisation of the gold price in some shape or form, or else are using their gold operations to raise foreign currencies for exchange rate intervention and/or system liquidity. But both routes are aiming at the same outcome. i.e. stability of the financial system.
At the end of the day, the gold price has always been a barometer that central banks strive to keep a lid on and which they aim to stabilise or smoothen the gyrations of, given that the alternative – a freely formed and unmanipulated gold price – would thwart their coordination of fiat currency exchange rates, interest rates and inflation targets.
Interestingly, in addition to the new spreadsheet of gold holdings data, the Bank of England gold web page now includes a link to a new 1 page ‘Gold Policy’ pdf document, which, looking at the pdf document’s properties, was only created on 30 January 2017. This document therefore also looks like it was written in conjunction with the new gold vault data rollout.
The notion of central banks accessing the liquidity of the London Gold Market via bullion banks is further developed in this Gold Policy document also. The document is quite short and merely states the following:
“GOLD ACCOUNTS AT THE BANK OF ENGLAND
1. The Bank primarily offers gold accounts to central bank customers. This is to support financial stability by providing central banks with secure custody for their gold reserves and access to the liquidity of the London gold market (particularly given the Bank’s location).
2. To facilitate, either directly or indirectly, access for central banks to the liquidity of the London gold market, the Bank will also consider providing gold accounts to certain commercial firms. In deciding whether to provide an account, the Bank will be guided by the following criteria.
a. The firm’s day to day activities must support the liquidity of the London gold market. b. Specifically, the Bank may have regard to a number of factors including but not limited to: evidence of active or prospective trading with a central bank customer; or whether the firm has committed to honour buy and sell prices.
3. Access to a gold account remains at the sole discretion of the Bank.
4. The Bank will review this policy periodically.”
The Vault Data
Nick Laird has now produced a series of impressive charts of this new Bank of England data on his website GoldChartsRUS. Plotting the series of 72 months of gold holdings data over January 2011 to December 2016 yields the below chart.
On average, the Bank’s vaults held 5457 tonnes of gold over this 6 year period. The minimum amount of gold held was 4693 tonnes at the end of March 2016, while the maximum quantity of gold held was 6250 tonnes at the end of February 2013.
The overall trend in the chart is downward with a huge outflow of gold bars from the bank’s vaults from the end of February 2013 to the end of March 2016.
As of January 2011, the BoE held just over 5500 tonnes of gold bars in its vaults. Gold holdings rose until the end of August 2011 and peaked at nearly 5900 tonnes before falling to 5600 tonnes at year-end 2011. Overall in 2011, the holdings fluctuated in a 400 tonne range, trending up during the first 8 months, and down during the latter 4 months.
This downtrend only lasted until January 2012, at which point BoE gold holdings totalled about 5450 tonnes. For the remainder of 2012, BoE gold under custody rose sharply, reaching 6200 tonnes by the end of 2012, a level near the ultimate peak in this 6 year chart. The year 2012 was therefore a year of accumulation of gold bars at the Bank during which 750 tonnes were added.
The overall maximum peak was actually 6250 tonnes at the end of February 2013, after which a sustained downtrend evolved through the remainder of 2013. By December 2013, gold under custody at the Bank of England had fallen to 5670 tonnes, creating an overall outflow of 580 tonnes of gold bars during 2013.
The outflow of gold continued during 2014 with another 470 tonnes flowing out of the Bank, leading to end of year 2014 gold holdings of just 5200 tonnes. The outflow also continued all through 2015 with only 4780 tonnes of gold in custody at the end of December 2015. The Bank therefore lost another 440 tonnes of gold bars in 2015.
Overall, that makes an outflow of 1490 tonnes of gold from the Bank’s vaults over the 3 years from 2013 to 2015 inclusive. This downtrend lingered for 3 more months, with another 80 tonnes lost, which brought the end of March 2016 and end of April 2016 figures to a level of about 4700 tonnes, which is the overall trough on the chart. It also means that there was a net outflow of 1570 tonnes of gold bars from the Bank’s vaults from the end of February 2013 to the end of March / April 2016.
A new uptrend / inflow trend began at the end of April 2016 and continued to the end of November 2016, where gold custody holdings peaked again at about 5123 tonnes before levelling off at the end of December 2016 at 5102 tonnes. Therefore, from the end of April 2016 to the end of December 2016, the Bank of England vaults added 400 tonnes of gold bars.
The gold holdings of the vast majority of central banks have remained stagnant over the 2011 – 2016 period, the exceptions being the central banks of China and Russia. But Russia buys domestically mined gold and stores it in vaults in Moscow and St Petersburg, so this would not affect gold holdings at the Bank of England. China’s central bank, the People’s Bank of China (PBoC), is known to buy its gold on the international market, including the London Gold Market. It then monetizes this gold (classifies it as monetary gold), and airlifts it back to China. But these Chinese purchases don’t show up in UK gold exports because monetary gold is exempt from trade statistics reporting. However, if China was surreptitiously buying gold from other central banks with gold accounts at the Bank of England or buying gold from bullion banks with gold accounts at the BoE, then some of the gold outflows from the BoE could be PBoC gold purchases. But without central bank specific data, its difficult to know.
But what is probably true is that the fluctuations in the quantity of gold stored in the Bank of England vaults are more do to with the gold holdings of bullion banks and less to do with the gold holdings of central banks, for the simple reason that central bank gold holdings are relatively static, or the least the central banks claim that their gold holdings are static. This does not take into account the gold lending market which the central banks and bullion banks go to great lengths to keep secret.
There is also a noticeable positive correlation between the movement of the US Dollar gold price and the inflows/outflows of gold to and from the Bank of England vaults, as the above chart demonstrates.
Bullion Bank gold accounts at the BoE
One basic piece of information that the Bank of England’s new vault storage data lacks is an indication of how many central banks and how many commercial banks are represented in the data.
In its first quarterly report from Q1 2014,the Bank of England states that 72 central banks operate gold accounts at the bank of England, a figure which includes a few official sector organisations such as the International Monetary Fund (IMF), European Central Bank (ECB), and Bank for International Settlements (BIS). This number would not have changed much in the meantime, so we can assume that the gold holdings of about 72 central banks are represented in the new data. But the number of commercial banks holding gold accounts at the Bank of England is less clear-cut.
The 5 gold clearing banks of the LPMCL all hold gold accounts at the Bank of England. Why? Because it says so on the LPMCL website:
“Each member of LPMCL has vaulting facilities under its control for the storage of gold and/or silver, plus in the case of gold bullion, account facilities at the Bank of England, which have contributed to the development of bullion clearing in London.”
The LPMCL also states that its clearing statistics include:
“Transfers over LPMCL Clearing Members’ accounts at the Bank of England.”
Additionally, the LPMCL website states that their
“clearing and vaulting services help facilitate physical precious metal movement logistics, location swaps, quality swaps and liquidity management.”
The Bank of England’s reference in its new ‘Gold Policy’ document to commercial banks needing to be “committed to honour buy and sell prices” is a reference to market makersand would cover all 13 LBMA market makers in gold, which are the 5 LPMCL members and also BNP Paribas, Citibank, Goldman Sachs, Merrill Lynch, Morgan Stanley, Société Générale, Standard Chartered Bank, Toronto-Dominion Bank. But there are also gold trading banks that make a market in gold which are not officially LBMA market makers, such as Commerzbank in Luxembourg which claims to be one of the biggest bullion banks in the world.
So I would say that lots of other bullion banks (of which there about 40 in total) have gold accounts at the Bank of England in addition to the 13 official LBMA market makers.
More fundamentally, any bullion bank that is engaged in gold lending with central banks (the central banks being the lenders and the bullion banks being the borrowers) would need a gold account at the Bank of England. I counted 28 bullion banks that have been involved with borrowing the gold of just one central bank, the central bank of Bolivia (Banco Central de Bolivia – BCB) between 1998 and 2016. Some of these banks have since merged or exited precious metals trading, but still, it gives an estimate of the number of bullion banks that have been involved in the gold lending market. The Banco Central de Bolivia’s gold lending activities will be covered in some forthcoming blog posts.
Bullion banks that are Authorised Participants (APs) for gold-backed ETFs such as the SPDR Gold Trust (GLD) or iShares Gold Trust (IAU) may also have gold accounts at the Bank of England. I say may have, because in practice the APs leave it up to the custodians such as HSBC and JP Morgan to allocate or deallocate the actual physical gold flowing in and out of the ETFs, but HSBC on occasion uses the Bank of England as a sub-custodian for GLD gold (see “SPDR Gold Trust gold bars at the Bank of England vaults” for details), so if some of the APs want to keep their own stash of allocated physical gold in relation to ETF trading, it would make sense for them to have a gold account at the Bank of England.
As to how much gold the GLD stores at the Bank of England and how regularly this occurs is still opaque because the SEC does not require the GLD filings to be very granular, however there is a very close correlation between inflows and outflows from GLD and the inflows and outflows from the Bank of England vaults, as the following chart clearly illustrates.
As gold was extracted from the GLD beginning in late 2012, a few months later the Bank of England gold holdings began to shrink also. This trend continues all the way through 2013, 2014 and 2015. Then as the amount of gold began to increase in the GLD at the end of 2015, the gold holdings at the Bank of England began to increase also. Could this be bullion banks extracting gold from the GLD, then holding this gold at the Bank of England and then subsequently exporting it out of the UK?
Some of it could, but UK gold net exports figures suggest that gold was withdrawn from both the Bank of England vaults and from the ETF gold stored at commercial gold vaults (run by HSBC and JP Morgan), after which it was exported.
Looking at the above chart which plots Bank of England gold holdings and UK gold imports and exports (and net exports) is revealing. As Nick Laird points out in this chart, over the 2013 to 2015 period during which the Bank of England gold holdings fell by 1500 tonnes, there were UK net gold export flows of 2500 tonnes, i.e. 2500 tonnes of gold flowed out of London gold vaults, so an additional 1000 tonnes had to come from somewhere apart from the Bank of England vaults.
The new monthly vault holdings data from the Bank of England can now also be compared to the amount of gold reported by the Bank of England in its annual reports. The figures the Bank reports in the annual report are as of the end of February. These figures are only reported in Pounds Sterling, not quantities, so they need to be either converted to USD and divided by the USD LBMA Gold Price on the last day of February, or else just divided by the GBP LBMA Gold Price on that day.
For end of February 2015, the calculated total for gold held at the Bank of England (based on the annual report) came out at 5,134 tonnes. Now the Bank of England data says 5126 tonnes which is very close to the calculation. For February 2016, the calculation came out at 4725 tonnes. The new Bank of England data now says 4730 tonnes, so that’s pretty close also.
This new Bank of England data is welcome and the Bank of England has taken a step towards greater transparency. However, it would be more useful if the Bank published a breakdown of how much of this gold is held by central banks and how much is held by bullion banks, along with the number of central banks and number of bullion banks that the data represents. Two distinct sets of data would be ideal, one for central bank custody holdings and the other for bullion bank custody holdings. The Bank most likely would never publish two sets of data as it would show bullion bank gold storage activity for the whole world to see.
While the Bank of England has now followed through with its promise to publish its gold vault holdings, the LBMA has still not published gold vault data for the commercial gold vault providers, i.e. its members HSBC, JP Morgan, ICBC Standard Bank, Brinks, Malca Amit, Loomis and G4S. Where is this data, why is there a delay, and why has it not yet been published?
As a reminder, the Financial Times article in early February said that the LBMA would publish gold vault holdings data that would:
“show gold bars held by the BoE, the gold clearing banks, and those [vaults] operated by the security companies such as Brink’s”
The Financial Times article also said that:
“HSBC and JPMorgan, London’s biggest bullion banks, are backing the initiatives by the LBMA to improve transparency.”
With the gold holdings data on the other London vaults still not published, it begs the question, has there been a change of mind by HSBC and JP Morgan, two of the LBMA’s largest and most powerful members?
“Reputedly [the Bank of England vaults are] the second largest vault in the world with approximately 500,000 gold bars held in safe custody on behalf of its customers, including LBMA members, central banks, international financial institutions and Her Majesty’s Treasury.”
A holding of 500,000 Good Delivery gold bars is equal to 6250 tonnes. However, according to the Bank of England’s own figure for month end December 2016, the Bank of England only holds 5100 tonnes of gold in custody (408,000 Good delivery gold bars). Therefore, the LBMA is overstating the Bank of England’s holdings by 1150 tonnes, unless, and it’s unlikely, that the BoE vaults have seen huge gold bar inflows in the last 4 months.
In a bizarre series of events that have had limited coverage but which are sure to have far-reaching consequences for benchmark pricing in the precious metals markets, the LBMA Gold Price and LBMA Silver Price auctions both experienced embarrassing trading glitches over consecutive trading days on Monday 10 April and Tuesday 11 April. At the outset, its worth remembering that both of these London-based benchmarks are Regulated Benchmarks, regulated by the UK’s Financial Conduct Authority (FCA).
In both cases, the trading glitches had real impact on the benchmark prices being derived in the respective auctions, with the auction prices deviating noticeably from the respective spot prices during the auctions. It’s also worth remembering that the LBMA Gold Price and LBMA Silver Price reference prices that are ‘discovered’ each day in the daily auctions are used to value everything from gold-backed and silver-backed Exchange Traded Funds (ETFs) to precious metals interest rate swaps, and are also used widely as reference prices by thousands of precious metals market participants, such as wholesalers, refineries, and bullion retailers, to value their own bi-lateral transactions.
Although the gold and silver auctions are separately administered, they both suffer from limited direct participation due to the LBMA only authorising a handful of banks to directly take part. Only 7 banks are allowed to participate directly in the Silver auction while the gold auction is only currently open to 14 entities, all of which are banks. Limited participation can in theory cause a lack of trading liquidity. Added to the mix, a central clearing option was introduced to the LBMA Gold Price auction on Monday 10 April, a day before Tuesday’s gold auction screw-up. The introduction of this central clearing process change saw four of the direct participants suspended from the auction since they had not made the necessary system changes in time to process central clearing. This in itself could have caused a drop in liquidity within Tuesday’s gold auction as it reduced the number of possible participants.
Other theories have been put forward to explain the price divergences, such as the banks being unwilling to hedge or arbitrage auction trades due to the advent of more stringent regulatory changes to prevent price manipulation. While this may sound logical in theory, no one, as far as I know, has presented empirical trade evidence to back up this theory. There is also the possibility of deliberate price manipulation of the auction prices by a participant(s) or their clients, a scenario that needs to be addressed and either ruled out or confirmed.
ICE Benchmark Administration (IBA), the administrator of the LBMA Gold Price, also introduced a price calculation Algorithm into the gold auction in mid-March 2017, a change which should also be considered by those seeking to find a valid explanation for the gold auction price divergence where the opening price kept falling through multiple auctions rounds whilst the spot price remained far higher. Could the algorithm have screwed up on 11 April?
Whatever the explanations for the price divergences, these incidents again raise the question as to whether these particular precious metals auctions are fit for purpose, and why they were designed (and allowed to be designed) at the outset to explicitly block direct participation by nearly every precious metals trading entity on the planet except for a limited number of London-based bullion bank members of the LBMA.
LBMA Silver Price fiasco
First up, on Monday 10 April, buried at the end of a Reuters News precious metals market daily news wrap was a very brief snippet of news referring to an incident which dogged the LBMA Silver Price during Monday’s daily auction (an auction which starts at midday London time). According to Reuters:
“silver prices slipped after the LBMA silver price benchmark auction was paused for 17 minutes after a circuit breaker was triggered when the auction price moved outside of the spot range, the CME said in a statement.”
What exactly the CME meant is unclear because whatever statement Reuters was referring to has not been released on the CME Group website or elsewhere, and Reuters did not write a separate news article about the incident.
To recap, the LBMA Silver Price is administered by Thomson Reuters on a calculation platform run by the CME Group, and operated on a contract basis on behalf of the London Bullion Market Association (LBMA). However, there is nothing anywhere on the CME’s LBMA Silver Price web page, or on the Thomson Reuters LBMA Silver Price web page, or on the LBMA website, in the form of a statement, comment or otherwise, referring to this ‘circuit breaker’ that persisted for ’17 minutes’ in the LBMA Silver Price auctionduring which time the ‘auction price moved outside of the spot range‘
On its calculation platform, CME makes use of a pricing algorithm to automatically calculate a price for each round of the LBMA Silver Price auction (excluding the first auction round). From page 8 of its LBMA Silver Price Methodology Guide:
“3.7 Starting Price
The initial auction price value is determined by the auction platform operator by comparing multiple Market Data sources prior to the auction opening to form a consensus price based on the individual sources of Market Data. The auction platform operator enters the initial auction price before the first round of the auction begins….”
“3.4 End of Round Comparison
If the difference between the total buy and sell quantity is greater than the tolerance value, the auction platform determines that the auction is not balanced, automatically cancels orders entered in the auction round by all participants, calculates a new price, and starts a new round with the new price.”
There is also a manual price override facility which can be invoked if needed:
3.8 Manual Price Override
In exceptional circumstances, CME Benchmark Europe Ltd can overrule the automated new price of the next auction round in cases when more significant or finer changes are required. When doing so, the auction platform operator will refer to a composition of live Market Data sources while the auction is in progress.”
As to why the “auction platform operator” did not invoke these manual override powers and seek market data sources during the time in which the silver auction was ‘stuck’ for 17 minutes is unclear. A 17 minute pause would presumably be, in the CME’s words, ‘exceptional circumstances’.
Unfortunately, neither the CME website, the Thomson Reuters website, or the LBMA website provides intra-round pricing data for the LBMA Silver Price, so anyone who doesn’t have a subscription to the live data of the auction is well and truly left in the dark as to what actually happened on Monday 10 April. Unlike the LBMA Gold Price auction which at least provides an ‘Auction Transparency Report’ for each auction (see below), the LBMA Silver Price auction is sorely lacking in any public transparency whatsoever.
But what is clear from the Reuters information snippet is that the LBMA Silver Price auction on Monday 10 April suffered a serious trading glitch, that saw the prices that were being formed in the auction deviate from where the silver spot price was trading during that time. This price deviation suggests a lack of trading liquidity in the auction and/or an inability of the participants to hedge their trades in other trading venues. As to whether the final LBMA Silver Price that was derived and published as the daily benchmark price on 10 March was outside the spot range (and above or below spot) is not mentioned in the Reuters report.
The complete opacity about this incident is concerning but not really surprising since nearly everything in the London precious metals markets is shrouded in secrecy, and corporate communication in this area is truly abysmal.
Recalling that Thomson Reuters and CME announced in early March that they are abruptly pulling out of the contract for administrating and calculating the LBMA Silver Price, this latest fiasco is unwelcome news for the LBMA – CME – Thomson Reuters triumvirate, and raises further questions for the FCA as to whether this Silver auction and benchmark should even be allowed to continue in its present or similar form.
LBMA Gold Price fiasco
Turning to the London gold auction, on the afternoon of Tuesday 11 April, the LBMA Gold Price auction (which starts at 3:00pm London time) experienced what can only be described as a shocking and serious trading fiasco which has real world consequences for all trading entities that use the LBMA Gold Price Benchmark reference price (and there are many that do so). As a reminder, ICE benchmark Administration (IBA) administers the daily LBMA Gold Price auctions on behalf of the LBMA.
“London’s gold price benchmark fixed some $12 below the spot price on Tuesday afternoon as the auction appeared to become locked in a downward spiral. From an initial $1,265.75, close to the spot price at the time, the auction price ratcheted steadily lower before fixing at $1,252.90 in the ninth round. From the fifth round to the eighth the bid and offer volumes remained frozen, unable to match.“
“This came a day after ICE introduced clearing for the LBMA Gold Price auction”
Reuters concludes its article by noting that the ICE clearing was introduced:“before several participating banks had the necessary systems in place.”
“As a result, China Construction Bank, Societe Generale, Standard Chartered and UBS are yet to confirm a date for their participation in the cleared auction.. ICE declined to comment. The LBMA, which owns the intellectual property rights to the auction, was not immediately available to comment.”
This forced reduction in the number of participants in the auction seems to be relevant to the issue and therefore requires further scrutiny.
ICE Central Clearing – Foisted on the LBMA Gold Price auction?
In mid-October 2016 during the LBMA precious metals conference in Singapore, ICE Benchmark Administration announced that it would introduce central clearing into the London Gold Price by utilizing a series of daily futures contracts which it planned to launch in February 2017. The introduction of central clearing into the auction was initially planned for March 2017.
“IBA gave a central clearing update to the Committee, notifying them that the cleared instrument would be launched in January 2017 and the auction trades could be routed there from March 2017. The Committee were informed that IBA had spoken to every bank and every bank wanted to move. Discussion moved to the technical implications for this new model and IBA’s primary wish to keep running a healthy auction.”
“From March 2017, subject to regulatory review, centrally cleared settlement will be available for transactions which originate from IBA’s gold auction underlying the LBMA Gold Price.
This will give firms the choice of settling their trades bilaterally against each counterparty (as they currently do), or submitting their trades to clearing and settling versus the clearing house. This mechanism removes the requirement for firms to have bilateral credit lines in place with all of the other Direct Participants in the auction.
Central clearing opens the auction to a broader cross-section of the market. It also facilitates greater volume in the auction.“
By the end of March 2017, the above statement had been altered from March 2017 to “Q2 2017″ with ICE pushing back the launch date for the introduction of central clearing:
“From Q2 2017, subject to regulatory review, centrally cleared settlement will be available for transactions which originate from IBAs gold auction underlying the LBMA Gold Price….”
Reuters again covered these ICE clearing delays in a series of articles during March, highlighting the fact that 4 of the 13 banks that are direct participants in the LBMA Gold Price auction were not ready for the introduction of central clearing due to delays in making unspecified changes to their internal IT systems that would allow such central clearing processing. So anybody who had been reading these Reuters articles would have been aware that there were risks on the horizon in terms of some of the LBMA Gold Price auction participants being slow in being ready for the changes.
“U.S.-based exchange operator ICE has already pushed back the launch of its service by several weeks to allow the banks and brokers who participate in the auction to adapt their IT systems, four sources with direct knowledge of the matter told Reuters.”
“Sources at many participant banks said that they were unhappy with the speed at which ICE was seeking to introduce clearing, which require investment in IT processes and back office systems and raise complex compliance issues.”
“However, at least four of the 14 banks and brokers who participate in the LBMA Gold Price auction will still not be ready to use the new system.
Banks that are not ready would be suspended from the auction until they have the necessary IT infrastructure in place or would have to participate through other players who could clear deals, according to the sources.
ICE’s readiness to provoke such disruption illustrates how much it wants to avoid further delays that could torpedo its ambitions to become the dominant exchange in London’s vast bullion market, market sources said”
“two sources told Reuters that ICE had again delayed and there was now no set start date.”
“Sources earlier told Reuters that Societe Generale, Standard Chartered, ICBC Standard Bank and China Construction Bank would not be ready to clear the LBMA auction in time for April 3.”
Again interestingly, ICE’s desire to promote its own gold futures contracts was seen as a primary driver for trying to rush through the introduction of central clearing for the gold auction, as doing so would add volume to ICE’s daily gold futures contracts:
“market sources say ICE plans to use clearing of the LBMA Gold Price auction, which it administers, to funnel business to its contracts and give it a head start over rivals.”
As a reminder, ICE and CME have both recently launched gold futures contracts connected to the London market, and the London Metal Exchange (LME) plans to launch its own suite of London gold futures contracts in early June.
Central clearing uses exchange for physical (EFP) transactions in the daily futures contracts which are then cleared at ICE Clear US. The futures have daily settlement each day between 3:00 pm and 3:05 pm London time. But how the whole process ties together is still quite puzzling. An email to the IBA CEO asking for details of how the futures are linked to the auction went unanswered.
So what was this downward spiral that the LBMA Gold Price auction experienced on the afternoon of Tuesday 11 April when it became, in the words of Reuters, locked in a downward spiral?
Let’s look at the ICE Auction Transparency Reports for the few days before and during the 11 April afternoon fiasco. These reports show the number of auction rounds, the number of participants,and the bid and offer volumes for each round as well as the price at the end of each round.
Fourteen entities are now authorized to be direct participants in the LBMA Gold Price auction, 13 of which are banks, the other being new participant INTL FCStone since early April. INTL FCStone is a financial services company that has a slant towards commodities. The 13 banks are:
Bank of China
Bank of Communications
China Construction Bank
Industrial and Commercial Bank of China (ICBC)
HSBC Bank USA
JPMorgan Chase Bank (London Branch)
The Bank of Nova Scotia – ScotiaMocatta
Unlike the old London Gold Fixing which had 5 member banks that were obliged to always turn up (and since 2004 dial in) for every auction, this LBMA Gold Price auction does not require all the authorized participants to dial-in. Most of the time, far fewer than the full contingent turn up. For example on Friday 7 April, 8 banks turned up at the morning auction while only 7 banks turned up at the afternoon auction (i.e only a 50% turnout). However, Friday 7 April is also relevant since that was the last day before ICE introduced central clearing to the gold auction.
Fast forwarding to the morning gold auction on Monday 10 April when ICE first introduced central clearing, you can see from the below auction report that only 5 banks participated. This is the same small number that took part in the former London Gold Fixing which was run by the infamous and scandal ridden London Gold Market Fixing Limited and which consisted of Deutsche Bank, Barclays, HSBC, Scotiabank and Société Générale.
The reason the turnouts after the introduction of central clearing are so low is that 4 of the direct participant banks have been excluded from the auction due to not being ready to implement central clearing – a fact predicted by Reuters News in March. This means that the usual number of between 7-10 banks participating in the auction has now been reduced by 4, as four banks cannot take part. As Reuters said on 21 March “Banks that are not ready would be suspended from the auction until they have the necessary IT infrastructure in place”.
The irony of this debacle is that the participating banks all already have bilateral credit limits with each other and so don’t need to do central clearing in the auction. Only new /future direct participants which do not have bilateral credit lines technically need to utilize the clearing solution.
Central clearing is supposed to make it easier for a far wider range and number of participants to take part. But if this entails enhancements to IT systems that some of the most sophisticated investment banks on the planet are struggling with, what hope is there for other precious metals trading entities to participate.
But some reason – probably to try to kickstart the trading volume in its daily gold futures contracts – ICE has made it mandatory for all existing direct participants (the bullion banks) to open clearing accounts and get their IT systems in shape to use clearing.
“Central clearing for the auction is enabled by effecting Exchange for Physical (“EFP”) transactions into the new physically settled, loco London gold daily futures contract which is traded on ICE Futures U.S. The EFPs establish positions in the futures contract which are cleared and can be physically delivered at ICE Clear U.S“
and Direct participants (DPs) “must establish a clearing account with an ICE Clear U.S. Clearing member” so as to be able to use this account to clear auction trades.
However, “DPs may still maintain credit lines to settle bilaterally against other DPs” and “DPs can elect, for each counterparty, to clear or settle their auction transactions bilaterally.” If this is so, then why the need to force these banks to open a clearing account and push through complex IT changes?
The ICE LBMA Gold Price web page now includes a double asterisk next to the names of the culprit banks that are not ready for central clearing. These banks are China Construction Bank, Société Générale, Standard Chartered, and UBS. the double asterisk states that “** Date of participating in the cleared auction to be determined.”
So now, more than 2 years after the LBMA Gold Price has been introduced, we are back to a situation where only 5 large bullion banks are participating in a daily gold price auction, an auction which has huge ramifications for the reference pricing of gold across myriad gold markets around the world.
Both of the auctions on 10 April finished within the first round, with buy volume and sell volume in balance, so there was no need for subsequent auction rounds.
Turning to the morning auction of Tuesday 11 April, only a measly 4 banks took part in the first round of the auction, and 5 participants took part in rounds 2 and 3. The bid and ask volumes were not that much out of balance, and the auction finished after 3 rounds.
Turning to the afternoon auction of 11 April, the price action commentary provided by Reuters was as follows:
“from an initial $1,265.75, close to the spot price at the time, the auction price ratcheted steadily lower before fixing at $1,252.90 in the ninth round. From the fifth round to the eighth the bid and offer volumes remained frozen, unable to match.“
Below you can see visually see what happened round by round from the first round price of $1,265.75 where there was zero bid volume and 125,217 ozs (nearly 4 tonnes) of ask volume, through the fifth to (actually) the ninth rounds where bid volume was an unchanging 92,873 ozs and ask volume was an unchanging 107,090 ozs, but still the price fell from $1,260.50 to fix in round 9 at $1,252.90, i,e, the price fell $7.60 in 2 minutes while the volumes didn’t budge. And most critically, the fixing price was $1252.90 while the spot price was trading at $1267 at that time.
“the benchmark ended up being set almost $15 dollars below where spot prices were trading at the time. The PM Gold Price showed a benchmark at $1,252.90 an ounce; however at the time, spot gold prices were trading around $1,267 an ounce, with prices heading towards a new five-month high.”
How could this happen? How could the auction price diverge so much from the spot price at that time and how could the auction go through round after round lowering the price while the bid and ask volumes did not change and while the spot price was actually far higher than any of the prices in the auction?
Kitco’s explanation, which is mostly based on the view of one person, Jeff Christian of the CPM Group, put the problem down to “poorly conceived regulations and a faulty price discovery mechanism“, i.e. a lack of liquidity due to banks being scared off by tightening regulations, and that this “sharp reduction in liquidity during the auction process” is causing “a large discrepancy in prices“. Christian also said that “because of regulations, banks and other financial institutions are backing away from becoming market makers.”
But this reasoning of backing away due to regulations is not backed up by the facts for the simple reason that banks have continued to join the LBMA Gold Price auction at a rapid rate over the last 2 years, i.e. there is a trend of ever more banks applying to be authorized to participate in the auction. For example, since the auction was launched on 20 March 2015 with 6 banks, 9 more banks have signed up JP Morgan, Morgan Stanley, Standard Chartered, Bank of China, ICBC, China Construction Bank, Bank of Communications, Toronto Dominion Bank, and INTL FCStone. Note that Barclays was one of the original six banks in the auction but dropped out after it downscaled its the precious metals business in London. There are also the same number of LBMA Market Makers now as there were two years ago, in both cases 13 LBMA Market Makers.
Kitco’s article also fails to mention the central clearing implementation fiasco brought about by ICE’s rush to channel activity into its gold futures contracts and Kitco even fails to realize that 4 banks were suspended from the auction due to this central clearing issue.
Another factor relevant to the screwed up afternoon auction on 11 April that should be considered is the fact that in mid-March 2017, ICE Benchmark Administration introduced a price algorithm into the LBMA Gold Price auction. This fact has been totally ignored by the financial media.
From a human Chairperson to an automated Algorithm
Up until mid March 2017, the LBMA Gold Price auction used a human ‘independent chairperson’ to choose the opening price in the auction and also the auction price in each subsequent round. The identities of these independent chairpersons have never been divulged by ICE nor the LBMA.
Critically, sometime during the 3rd week of March 2017, ICE Benchmark Administration (IBA) introduced a pricing algorithm into the LBMA Gold Price auction. This change in procedure (moving from an auction chairperson to an auction pricing algorithm) was not actively highlighted by either ICE or the LBMA but is clear from looking at Internet Archive imprints of the ICE LBMA Gold Price webpage.
“The auction process has an independent chairperson, appointed by IBA to determine the price for each round and ensure that the price responds appropriately to market conditions.”
See screenshot below for the same statement – taken from the same webpage:
Bullet point 1 of the Auction Process for the 9 March version of the webpage also refers to the chairperson as being responsible for setting the starting price and the price of each subsequent round “in line with current market conditions and the activity in the auction.”
But by 16 March, when the next imprint of the LBMA Gold Price page was made by the Internet Archive, the reference in the methodology section to an independent chairperson had been fully deleted, and bullet point 1 had been changed from mentioning a chairperson to discussing an algorithm, specifically changed to “IBA sets the starting price and the price for each round using an algorithm that takes into account current market conditions and the activity in the auction.”
See screenshot below for the same statement – taken from the same webpage:
So if there is an algorithm that is taking into account current market conditions in addition to activity in the auction, why did this algorithm not take the current spot prices into account over rounds 4 – 9 of the LBMA Gold Price auction on the afternoon of Tuesday 11 April?
Furthermore, for such a major change to the methodology and auction process in an auction whose benchmark price is widely used in the gold world, it’s very surprising that neither ICE, nor the LBMA, nor the London financial media mentioned this substantial algorithmic change.
In early December 2016, ICE published an LBMA GOLD PRICE Methodology Consultation in which one of the consultation’s proposed changes was “the introduction of an algorithm to determine the price for each auction round“.
The December 2016 document noted that:
“IBA’s auction process is currently that the auction chair sets the price for each Round in line with current market conditions and the activity in the auction”
“IBA currently has a panel of auction chairs who are independent of any firm associated with the auction, including Direct Participants. The chairs are externally sourced but work with IBA to deliver a robust process for determination of the LBMA Gold Price.
The chairs use their extensive market experience to set the round prices based on a pricing framework agreed with IBA. IBA chose to operate the auction using human chairs to make sure that the price could respond appropriately to market conditions from the outset.
IBA’s feedback from the market was that, at least in the early stages, the professional judgement of a human chairman was needed.“
“After operating the auction for more than a year, IBA started to develop an algorithm to set the auction’s starting price and subsequent round prices. IBA has now been testing and refining the algorithm over a number of months“
As per the proposal, the algorithm would replace the human chair, after which:
“Each auction will continue to be supervised by IBA’s analysts, and, if for any reason an auction did not progress as expected, IBA’s existing safeguards would be deployed to protect the integrity of the auction and the LBMA Gold Price benchmark“
These safeguards were stated as being three, namely:
- Pause the auction and restart, to give Participants an opportunity to contact clients or re-evaluate their positions
- Increase the imbalance threshold, if it appears that the auction will otherwise not finish
- Cancel an order, if it is compromising the integrity of the process and the relevant participant cannot be reached.
The proposals were pencilled in for implementation in Quarter 1, 2017.
Following the consultation, a “Methodology Consultation Feedback” document was published on the ICE Benchmark Administration website. One feedback respondent was concerned about who would be overseeing the daily auctions in the absence of a human chairperson, to which ICE answered:
“IBA can confirm that the auction will always be supervised by at least two IBA analysts. This approach is consistent with how we operate our other benchmarks.
Our aim is to put the auction on auto-pilot, not to make it driverless.
Unfortunately, from the wider gold market’s perspective, the LBMA Gold Price auction on the afternoon of Tuesday 11 April does indeed appear to have been ‘driverless‘ as it “did not progress as expected“, so it is now up to the LBMA and ICE to establish what the ‘IBA analysts’ were up to behind the driving wheel that day.
On its website, ICE states that the LBMA Gold Price methodology is “reviewed by the LBMA Gold Price Oversight Committee as documented in its Terms of Reference.” This Oversight Committee should also explain to the gold world what actually happened on the afternoon of 11 April.
Additionally, I find no explanation on ICE’s LBMA Gold Price webpage as to how exactly the automated algorithm works, what its logic rules are, how it was programmed etc.
The trading glitch with the LBMA Silver Price on Monday 10 April seems to have been completely missed by London’s financial media except for the brief reference by Reuters. The fact that there is no information on the CME, Thomson Reuters and LBMA websites about the issue should raise concern for users of this benchmark and for the UK’s regulator, the FCA. In an ideal world, there should be a full ‘outage’ report published on each of the 3 websites explaining what happened, but this will not happen in the shadowy and secretive London Silver Market.
Perhaps the auction price divergence in the LBMA Silver Price stems from a lack of liquidity brought on by the limited presence of auction participants, or due to the inability or unwillingness of participants to hedge or arbitrage their auction trades against the London OTC spot or other trading venues? The simple thing to do would be for CME, Thomson Reuters and the LBMA to explain themselves since this would minimize guesswork and to provide global silver market entities with clarity. Anything short of a full explanation by the parties concerned is irresponsible.
For the LBMA Gold Price auction, ICE Benchmark Administration needs to release a full ‘outage’ report and explanation on what exactly happened in the afternoon auction on 11 April and explain to the global gold market whether the introduction of central clearing was in any way responsible for the price divergence, and whether there are any conflicts of interest in trying to get banks to use its daily gold futures contracts. While they are at it, ICE should fully explain how the recent introduction of a pricing algorithm impacts the gold auction and whether this too had an impact on the auction price entering a downward spiral.
As the LBMA Silver Price and LBMA Gold Price are both Regulated Benchmarks, the FCA regulator needs to step up to the plate and for once show that it is on the side of the users of these benchmarks and not the powerful London banks.
Both of these auctions require full transparency and ease of direct participation by the full spectrum of the world’s gold and silver trading entities. Currently, they fall far short of these goals.
On 5 February, the Financial Times of London (FT) featured a story revealing that the London Bullion Market Association (LBMA) plans to begin publishing data on the amount of real physical gold actually stored in the London precious metals vaulting network. The article titled “London gold traders to open vaults in transparency push” can be read here (accessible via FT subscription or via free monthly FT read limit).
This new LBMA ‘monthly vault data’ will, according to the FT’s sources, be published on a three-month lagged basis, and will:
“show gold bars held by the BoE, the gold clearing banks, and those [vaults] operated by the security companies such as Brink’s, which are also members of the LBMA.”
The shadowy source quoted in the FT article is attributed to “a person involved in setting up the programme”, but at the same time, although “the move [to publish the data] is being led by the LBMA“, the same LBMA ”declined to comment” for the FT story. This then has all the hallmarks of a typical authorised leak to the media so as to prepare the wider market for the data release.
On 16 February, the World Gold Council in its “Gold Investor, February 2017″ publication featured a focus box on the same gold vault topic in its “In the News” section on page 4, where it states:
“Enhanced transparency from the Bank of England
The Bank of England is, for the first time, publishing monthly data revealing the amount of gold it holds on behalf of other central banks.
As a leading custodian of gold, with one of the largest vaults in the world, the Bank of England’s decision is highly significant. Not only will it enhance the transparency of the Bank’s own gold operations; it will also support the drive towards greater transparency across the gold market.
The data reveals the total weight of gold held within the Bank of England’s vaults and includes five years of historical data.”
The Proposed Data
Based on these two announcements, it therefore looks like the gold vault data release will be a combined effort between the LBMA and the Bank of England, the blood brothers of the London Gold Market, with the Bank of England data being a subset of the overall LBMA data. While neither of the above pieces mention a release date for the first set of data, I understand that it will be this quarter, i.e. sometime before the end of March. On a 3 month lagged basis, the first lot of data would therefore probably cover month-end December 2016, because that would be a logical place to start the current dataset, rather than, for example, November 2016.
While the Bank of England data looks set to cover a 5 year historical period, there is no indication (from the FT article) that the wider LBMA vault data will do likewise. From the sparse information in the FT article, the LBMA data will “show gold bars held“. Does it mean number of gold bars, or combined weight of gold bars? What exactly it means, we will have to wait and see.
The Bank of England data will capture “total weight of gold held“. Notice that in the above World Gold Council piece it also states that the data will cover the amount of gold that the Bank of England “holds on behalf of other central banks.” There is no mention of the amount of gold that the Bank of England holds on behalf of commercial bullion banks.
Overall, this doesn’t exactly sound like it is “enhancing the transparency of the Bank’s own gold operations” as the World Gold Council puts it. Far from it. Enhancing the transparency of the Bank of England’s gold operations would require something along the lines of the following:
Identities of all central banks and official sector institutions (ECB / IMF / BIS / World Bank) holding active gold accounts at the Bank of England. Active gold accounts meaning non-zero balances
Identities of all commercial / bullion banks holding active gold accounts at the Bank of England
A percentage breakdown between the central bank gold held in the Bank of England vaults and the bullion bank gold held in the Bank of England vaults
An indicator for each gold account as to whether it is a set-aside earmarked custody account or whether it is a fine troy ounce balance account
Information for each central bank and official sector institution as to whether any of “its” gold is lent, swapped or repo’d
Information for the bullion bank gold accounts as to whether the gold recorded in those accounts is borrowed, sourced from swaps, sourced from repos, or otherwise held as collateral for loans
Information on the gold accounts of the 5 LPMCL clearing banks showing how much gold each of these institutions holds each month and whether the Bank of England supplies physical gold clearing balances to these banks
Information on when and how often the London-based gold-backed ETFs store gold at the Bank of England, not just using the Bank of England as sub-custodian, but also storage in their own names, i.e. does HSBC store gold in its own name at the Bank of England which is used to supply gold to the SPDR Gold Trust
Information on whether and how often the Bank of England intervenes into the London Gold Market and the LBMA Gold Price auctions so as to supply gold in price smoothing and price stabilisation operations in the way that the Bank of England’s Terry Smeeton seems to have been intervening into the London Gold Market in the 1980s
Information on the BIS gold holding and gold transactions settlements accounts at the Bank of England and the client sub-account details and central bank identities for these accounts
Information on gold location swaps between gold account holders at the Bank of England and gold accounts at the Federal Reserve Bank of New York, the Banque de France, and the Swiss National Bank, and BIS accounts in those locations
Gold for oil swaps and oil for gold swaps
Anything less is just not cricket and does not constitute transparency.
And its important to remember that any publication of gold vault data by the LBMA and Bank of England is not being done because the LBMA suddenly felt guilty, or suddenly had an epiphany on the road to Damascus, but, as the FT correctly points out:
“the LBMA, whose members include HSBC and JPMorgan, hopes to head off the challenge and persuade regulators that banks trading bullion should not have to face more onerous funding requirements.”
The Current Data
As a reminder, there is currently no official direct data published on the quantity of real physical gold bars held within the London gold vaulting system. This vaulting system comprises the vaults of eight vault operators (see below for list).
Once a year in its annual report, the Bank of England provides a Sterling (GBP) value of gold held by its gold custody customers, while the LBMA website states a relatively static total figure of “approximately 6,500 tonnes of gold held in London vaults” that it claims are in the vaults in its network. But beyond these figures, there is currently no official visibility into the quantity of London Good Delivery gold bars held in the London vaults. There are, various ways of estimating London gold vault data using the Bank of England annual figure and the LBMA figure together with Exchange Traded Fund gold holdings and central bank divulged gold holdings at the Bank of England.
The September 2015 estimates calculated that there were 6,256 tonnes of gold in total in the London vaults, with 5,134 tonnes at the Bank of England (as of end February 2015), and 1,122 tonnes in London “not at the Bank of England“, all of which was accounted for by gold-backed ETFs which store their gold in London. These calculations implied that there was nearly zero gold stored in London outside the Bank of England that was not accounted for by ETF holdings.
The “Tracking the gold held in London” estimates from September 2016 used a figure of 6,500 tonnes of gold in total in the London vaults, and showed that there were 4,725 tonnes inside the Bank of England vaults, of which about 3,800 tonnes was known to be held by central banks (and probably a lot of the remainder was held by central banks also) and that there were 1,775 tonnes of gold outside the Bank of England. The article also calculated that there were 1,679 tonnes of gold in the gold backed ETFs that store their gold in London, so again, there was very little gold in the London vault network that was not accounted for by ETFs and central bank gold.
The Vaults of London
Overall, there are 8 vault operators for gold within the LBMA vaulting network. These 8 vault operators are as follows:
The Bank of England
HSBC Bank plc
JP Morgan Chase
ICBC Standard Bank Plc
Malca-Amit Commodities Ltd
G4S Cash Solutions (UK) Limited
Loomis International (UK) Ltd
HSBC, JP Morgan and ICBC Standard are 3 of the London Gold Market’s clearing banks that form the private company London Precious Metals Clearing Limited (LPMCL). The other two member of LPMCL are Scotia Mocatta and UBS. Brink’s, Malca-Amit, G4S and Loomis are the aforementioned security companies. The LBMA website lists these operators, alongside their headquarters addresses.
Bizarrely, the FT article still parrots the LBMA’s spoon-fed line that the vaults are “in secret locations within the M25 orbital motorway”. But this is far from the truth. Many of the London vault locations are in the public domain as has been covered, for example, on this website, and the FT knows this:
It’s slightly disappointing that we spend time and effort informing the London financial media where some of the London gold vaults are, and then they continue to parrot the LBMA’s misleading “secret locations” line. I put this fake news down to a decision by the FT editors, who presumably have a stake in playing along with this charade so as not to rock the boat with the powerful investment banks that they are beholden to.
The FT also reminds us in its article that “last year a gold vault owned by Barclays, which can house $80bn of bullion, was bought by China’s ICBC Standard Bank.“
This Barclays vault in London was built by and is operated by Brink’s, and presumably after being taken over by ICBC Standard, it is still operated by Brink’s. Logistically then, this ICBC Standard vault is most likely within the Brink’s complex, a location which is also in the public domain, and which even hosts an assay office as was previously mentioned here over a year ago. The Barclays vault (operated by Brink’s) is even mentioned in a Brink’s letter to the SEC in February 2014, which can also be seen here -> Brinks letter to SEC February 2014.
Given the fact that there are eight sets of vaults in the London vault system (as overseen by various groups affiliated to the LBMA such as the LBMA Physical Committee, the LBMA Vault Managers Working Party, the gold clearers (London Precious Metals Clearing Limited), and even the LBMA Good Delivery List referees and staff, then one would expect that whatever monthly vault data that the LBMA or its affiliates publishes in the near future, will break out the gold bar holdings and have a distinct line item in the list for each vault operator such as:
HSBC – w tonnes
JP Morgan – x tonnes
ICBC Standard – y tonnes
Brink’s – z tonnes
At the LBMA conference in Singapore last October, there was talk that there were moves afoot for the Bank of England to begin publishing data on the custody gold it holds on a more regular basis. It was also mentioned that this data could be extended to include the commercial bank and security carrier vaults but that some of the interested parties were not in favour of the idea (perhaps the representative contingents of the powerful HSBC and JP Morgan). Whatever has happened in the meantime, it looks like some data will now be released in the near future covering all of the participating vaults. What this data will cover only time will tell, but more data than less is always welcome, and these data releases might also help show how near or how far we were with earlier estimates in trying to ascertain how much gold is in the London vaulting system that is not accounted for by ETF holding or central bank holdings.
Revealing the extent of the gold lending market in London is critical though, but this is sure to remain a well-kept secret, since the LBMA bullion banks and the Bank of England will surely not want the general market to have any clue as to which central banks don’t really have any gold while still claiming to have gold (the old gold and gold receivables trick), in other words, that there is serious double counting going on, and that some of the central bank gold has long gone out the door.
The European Central Bank (ECB), creator of the Euro, currently claims to hold 504.8 tonnes of gold reserves. These gold holdings are reflected on the ECB balance sheet and arose from transfers made to the ECB by Euro member national central banks, mainly in January 1999 at the birth of the Euro. As of the end of December 2015, these ECB gold reserves were valued on the ECB balance sheet at market prices and amounted to €15.79 billion.
The ECB very recently confirmed to BullionStar that its gold reserves are stored across 5 international locations. However, the ECB also confirmed that it does not physically audit its gold, nor will it divulge a bar list / weight list of these gold bar holdings.
Questions and Answers
BullionStar recently put a number of questions to the European Central Bank about the ECB’s gold holdings. The ECB Communications Directorate replied to these questions with answers that appear to include a number of facts about the ECB gold reserves which have not previously been published. The questions put to the ECB and its responses are listed below (underlining added):
Question 1: “The 2015 ECB Annual Report states that as at 31 December 2015, the ECB held 16,229,522 ounces of fine gold equivalent to 504.8 tonnes of gold. Given that the ECB gold holdings arose from transfers by the respective member central banks, could you confirm the storage locations in which this ECB gold is currently held (for example at the Bank of England etc), and the percentage breakdown of amount stored per storage location.”
ECB Response: “The gold of the ECB is located in London, Paris, Lisbon, New York and Rome. The ECB does not disclose its distribution over these places. The gold of the ECB is stored there because it was already stored there before ownership was transferred to the ECB and moving it was seen and is seen as too costly.“
Question 2: “Could you clarify as to how, if at all, this gold is audited, and whether it physical audited by the ECB or by a 3rd party?”
ECB Response: “The ECB has no physical audit of its gold bars. The gold bars that the ECB owns are individually identified and each year the ECB receives a detailed statement of these gold deposits. The central banks where the gold is stored are totally reliable.“
Question 3:“Finally, can the ECB supply a full weight list of the gold bars that comprise the 504.8 tonnes of gold referred to above?”
ECB Response: “The ECB does not disclose this information.“
London, New York, Paris, Rome, Lisbon
Given that some of the information shared by the ECB has arguably not been in the public record before, each of the 3 ECB answers above is worth further exploration.
In January 1999, when the Euro currency was created (Stage 3 of Economic and Monetary Union), each founding member national central bank (NCB) of the Euro transferred a quantity of foreign reserve assets to the ECB. Of these transfers, 85% was paid to the ECB in the form of US dollars and Japanese Yen, and 15% was paid to the ECB in the form of physical gold.
Initially in January 1999, central banks of 11 countries that joined the Euro made these transfers to the ECB, and subsequently the central banks of a further 8 countries that later joined the Euro also executed similar transfers to the ECB.
All of the foreign exchange and gold reserves that were transferred to and are owned by the ECB are managed in a decentralised manner by the national central banks that initiated the transfers. Essentially, each national central bank acts as an agent for the ECB and each NCB still manages that portion of reserves that it transferred to the ECB. This also applies to the transferred gold and means that the gold transferred to the ECB never physically moved anywhere, it just stayed where it had been when the transfers of ownership were made.
That is why, as the ECB response to Question 1 states: “The gold of the ECB is stored there because it was already stored there before ownership was transferred to the ECB”.
What is probably most interesting about the latest ECB statement is that it names 5 city locations over which the ECB’s gold is stored. The 5 gold storage locations stated by the ECB are London, New York, Paris, Rome and Lisbon. Since the gold transferred to the ECB in 1999 by the national central banks would have already been stored in central banks gold vaults, these 5 city locations undoubtedly refer to the gold vaults of:
the Bank of England
the Federal Reserve Bank of New York
the Banque de France
the Banca d’Italia
Banco de Portugal
The fact the ECB’s gold holdings are supposedly stored at these 5 locations can be explained as follows:
Between 4th and 7th January 1999, 11 central banks transferred a total of €39.469 billion in reserve assets to the ECB (in the form of gold, cash and securities). Of this total, 15% was in the form of gold, amounting to 24 million ounces of gold (747 tonnes of gold) which was valued at that time at €246.368 per fine ounce of gold, or €5.92 billion. The 85% transferred in the form of currencies comprised 90% US Dollars and 10% Japanese Yen. See pages 152 and 153 of ECB annual report 1999 for more details.
The 11 central banks that made the transfers to the ECB in January 1999 were the central banks of Belgium, Netherlands, Germany, France, Luxembourg, Italy, Ireland, Austria, Finland, Spain and Portugal. See Table 1 for details of these gold transfers, and the amount of gold transferred to ECB ownership by each central bank.
The value of reserves transferred to the ECB by each national central bank were based on a percentage formula called a ‘capital key’ which also determined how much each central bank subscribed to the founding capital of the ECB. This capital key was based on equally weighting the percentage of population and GDP each Euro founding member economy represented, therefore central banks such as Deutsche Bundesbank, Banque de France, and Banca d’Italia comprised the largest transfers, as can be see in Table 1. It also meant that these 3 central banks transferred the largest amounts of gold to the ECB, with the Bundesbank for example transferring 232 tonnes of gold to the ECB.
The Bundesbank gold transfer to the ECB in January 1999 took place at the Bank of England. The Bundesbank actually confirmed in its own published gold holdings spreadsheet that this transfer took place at the Bank of England. See spreadsheet Column 5 (BoE tonnes), Rows 1998 and 1999, where the Bundesbank gold holdings fell by 332 tonnes between 1998 and 1999 from 1,521 tonnes to 1,189 tonnes and also see Column 20 where gold lending rose from 149 tonnes to 249 tonnes. Therefore, between 1998 and 1999, 232 tonnes of gold was transferred from the Bundesbank gold account at the bank of England to the ECB account at the Bank of England, and 100 tonnes was added to the Bundesbank’s gold loans.
Paris and Rome
The Banque de France currently stores the majority (over 90%) of its gold reserves in its own vaults in Paris, so it it realistic to assume that when the Banque de France transferred 159 tonnes of gold to the ECB in January 1999, it did so using gold stored in the Banque de France vaults in Paris. Likewise, it is realistic to assume that the Banca d’Italia, which currently stores half of its gold reserves at its own vaults in Rome, transferred 141 gold stored in its Rome vaults to the ECB in 1999. This would explain the Paris and Rome gold holdings of the ECB. While a few ex French colony central banks are known to have historically stored gold with the Banque de France in Paris, none of the founding members of the Euro (apart from the Bundesbank) are on the record as having stored gold in Paris, at least not for a long time. The Banca d’Italia is not known for storing gold on behalf of other national central banks.
Lisbon and New York
The Banco de Portugal currently holds its gold reserves in Lisbon and also at the Bank of England, the Federal Reserve Bank of New York (FRBNY), and with the BIS. The ECB gold stored in Lisbon, Portugal most likely refers to the 18.2 tonnes of gold transferred by the Banco de Portugal to the ECB in January 1999, because a) that makes most sense, and b) the Banco de Portugal is not known as a contemporary gold custodian for other central banks.
Of the other 7 central banks that transferred gold to the ECB in January 1999, the central banks of Austria, Belgium and Ireland store most of their gold at the Bank of England so are the most likely candidates to have made gold transfers to the ECB at the Bank of England. See BullionStar blog “Central bank gold at the Bank of England” for more details of where central banks are known to store gold.
The Netherlands and Finland currently store some of their gold reserves at the Bank of England and at the Federal Reserve Bank of New York and probably also did so in 1998/99, so one or both of these banks could have made transfers to the ECB at the FRBNY. Another contender for transferring gold held at the FRBNY is the Spanish central bank since it historically was a holder of gold at the NYFED. It’s not clear where the central bank of Luxembourg held or holds gold but it’s not material since Luxembourg only transferred just over 1 tonne to the ECB in January 1999.
Greece and Later Euro members
Greece joined the Euro in January 2001 and upon joining it transferred 19.5 tonnes of gold to the ECB. Greece is known for storing some of its gold at the FRBNY and some at the Bank of England, so Greece too is a candidate for possibly transferring New York held gold to the ECB. In theory, the ECB’s New York held gold may not have even arisen from direct transfers from Euro member central banks but could be the result of a location swap. Without the national central banks or the ECB providing this information, we just don’t know for sure how the ECB’s New York gold holdings arose.
Another 7 countries joined the Euro after Greece. These countries were Slovenia on 1st January 2007, Malta and Cyprus 1st January 2008, Slovakia 1st January 2009, Estonia 1st January 2011, Latvia 1st January 2014, and Lithuania 1st January 2015. The majority of these central banks made gold transfers to the ECB at the Bank of England. In total these 7 central banks only transferred 9.4 tonnes of gold to the ECB, so their transfers are not really material to the ECB’s gold holdings.
These sales explain why the ECB currently only holds 504.8 tonnes of gold:
i.e. 766.9 t (including Greece) – 271.5 t sales + 9.4 t smaller member transfers = 504.8 t
The ECB does not provide, nor has ever provided, any information as to where the 271.5 tonnes of gold involved in these 2005-2009 sales was stored when it was sold. The fact that the ECB still claims to hold gold in Paris, Rome and Lisbon, as well as London and New York, suggests that at least some of the gold transferred by the Banque de France, Banca d’Italia and Banco de Portugal in 1999 is still held by the ECB.
If the ECB had sold all the gold originally transferred to it by all central banks other than France, Italy, Portugal and Germany, this would only amount to 197 tonnes, so another 74 tonnes would have been needed to make up the shortfall, which would probably have come from the ECB holdings at the Bank of England since that is where most potential central bank and bullion bank buyers hold gold accounts and where most gold is traded on the international market.
Even taking into account Greece’s 19.4 tonne gold transfer to the ECB in January 2001, and excluding the French, Italian, German and Portuguese transfers in 1999, the ECB’s 271.5 tonnes of gold sales would still have burned through all the smaller transfers and left a shortfall. So the ECB gold sales may have come from gold sourced from all of its 5 storage loacations.
It’s also possible that one or more of the original 11 central banks transferred gold to the ECB that was stored at a location entirely distinct from the 5 currently named locations, for example gold stored at the Swiss National Bank. If that particular gold was then sold over the 2005-2009 period, it would not get picked up in the current locations. It’s also possible that some or all of the 271.5 tonnes of gold sold by the ECB over 2005-2009 had been loaned out, and that the ‘sales’ were just a book squaring exercise in ‘selling’ gold which the lenders failed to return, with the loan transactions being cash-settled.
No Physical Audit of ECB Gold
Given that the Euro is the 2nd largest reserve currency in the world and the 2nd most traded currency in the world, the ECB’s gold and how that gold is accounted for is certainly a topic of interest. Although the ECB’s gold doesn’t directly back the Euro, it backs the balance sheet of the central bank that manages and administers the Euro, i.e. the ECB.
The valuation of gold on the ECB’s annual balance sheet also adds to unrecognised gains on gold in the ECB’s revaluation account. Given gold’s substantial price appreciation between 1999 and 2015, the ECB’s unrecognised gains on gold amount to €11.9 billion as of 31 December 2015.
It is therefore shocking, but not entirely surprising, that the ECB doesn’t perform a physical audit of its gold bars and has never done so since initiating ownership of this gold in 1999. Shocking because this lack of physical audit goes against even the most basic accounting conventions and fails to independently prove that the gold is where its claimed to be, but not surprising because the world of central banking and gold arrogantly ignores and bulldozes through all generally accepted accounting conventions. Geographically, 2 of the locations where the ECB claims to store a percentage of its gold are not even in the Eurozone (London and New York), and infamously, the Bundesbank is taking 7 years to repatriate a large portion of its gold from New York, so the New York storage location of ECB gold holdings should immediately raise a red flag. Furthermore, the UK is moving (slowly) towards Brexit and away from the EU.
Recall the response above from the ECB:
“The ECB has no physical audit of its gold bars. The gold bars that the ECB owns are individually identified and each year the ECB receives a detailed statement of these gold deposits. The central banks where the gold is stored are totally reliable.“
Imagine a physical-gold backed Exchange Traded Fund (ETF) such as the SPDR Gold Trust or iShares Gold Trust coming out with such a statement. They would be run out of town. References to ‘totally reliable’ are all very fine, but ‘totally reliable’ wouldn’t stand up in court during an ownership claim case, and assurances of ‘totally reliable’ are not enough, especially in the gold storage and auditing businesses.
The ECB is essentially saying that these ‘statements’ of its gold deposits that it receives from its storage custodians are all that is needed to for an “audit” since the custodians are ‘totally reliable‘.
This auditing of pieces of paper (statements) by the ECB also sounds very similar to how the Banca d’Italia and the Deutsche Bundesbank conduct their gold auditing on externally held gold i.e. they also merely read pieces of paper. Banca d’Italia audits “annual certificates issued by the central banks that act as the depositories” (the FRBNY, the Bank of England, and the SNB/BIS).
The Bundesbank does likewise for its externally held gold (it audits bits of paper), and solely relies on statements from custodians that hold its gold abroad. The Bundesbank actually got into a lot of heat over this procedure in 2012 from the German Federal Court of Auditors who criticised the Bundesbank’s blasé attitude and lack of physical auditing, criticism which the Bundesbank’s executive director Andreas Dombret hilariously and unsuccessfully tried to bury in a speech to the FRBNY in New York in November 2012 in which he called the controversy a “bizarre public discussion” and “a phantom debate on the safety of our gold reserves“, and ridiculously referred to the movies Die Hard with a Vengeance and Goldfinger, to wit:
“The days in which Hollywood Germans such as Gerd Fröbe, better known as Goldfinger, and East German terrorist Simon Gruber, masterminded gold heists in US vaults are long gone. Nobody can seriously imagine scenarios like these, which are reminiscent of a James Bond movie with Goldfinger playing the role of a US Fed accounting clerk.”
Where is the ECB Gold Bar Weight List?
Since, as the ECB states, it’s gold bars are “individually identified“, then gold bar weight lists of the ECB’s gold do indeed exist. This then begs the question, where are these weight lists, and why not release them if the ECB has nothing to hide?
Quickly, to define a weight list, a gold bar weight list is an itemised list of all the gold bars held within a holding which uniquely identifies each bar in the holding. In the wholesale gold market, such as the London Gold Market, the LBMA’s “Good Delivery Rules” address weight lists, and state that for each gold bar on a weight list, it must list the bar serial number, the refiner name, the gross weight of the bar, the gold purity of the bar and the fine weight of the bar. The LBMA also state that “year of manufacture is one of the required ‘marks’ on the bar”.
Recall from above that when the ECB was asked to provide a full weight list of its 504.8 tonnes of gold bars, it responded: “The ECB does not disclose this information.“
After receiving this response, BullionStar then asked in a followup question as to why the ECB doesn’t disclose a weight list of the gold bars. The ECB responded (underlining added):
“We would like to inform you that, while the total weight and value of the gold held by the European Central Bank (ECB) can be considered to be of interest to the public, the weight of each gold bar is a technicality that does not affect the economic characteristics of the ECB’s gold holdings. Therefore the latter does not warrant a publication.“
It is a very simple task to publish such a weight list in an automated fashion. The large gold backed ETFs publish such weight lists online each and every day, which run in to the hundreds of pages. Publication of a weight list by the ECB would be a very simple process and would prove that the claimed bars are actually allocated and audited.
The more evidence that is gathered about the refusal of central banks to issue industry standard gold bar weight lists, the more it becomes obvious that there is a coordinated understanding between central banks never to release this information into the public domain.
The most likely reason for this gold bar weight list secrecy is that knowledge of the contents of central bank gold bar weight lists could begin to provide some visibility into central bank gold operations such as gold lending, gold swaps, location swaps, undisclosed central bank gold sales, and importantly, foreign exchange and gold market interventions. This is because with weight list comparisons, gold bars from one central bank weight list could begin turning up in another central bank weight list or else turning up in the transparent gold holdings of vehicles such as gold-backed Exchange Traded Funds.
Instead of being fixated with the ECB’s continual disastrous and extended QE policy, perhaps some financial journalists could bring themselves to asking Mario Draghi some questions about the ECB gold reserves at the next ECB press briefing, questions such as the percentage split in storage distribution between the 5 ECB gold storage locations, why ECB gold is being held in New York, why is there no physical audit of the gold by the ECB, why does the ECB not publish a weight list of gold bar holdings, and do the ECB or its national central bank agents intervene into the gold market using ECB gold reserves.
The lackadaisical attitude of the ECB to its gold reserves by never physically auditing them is also a poor example to set for all 28 of the central bank members of the European System of Central Banks (ESCB), and doesn’t bode well for any ESCB member central bank in being any less secretive than the ECB headquarters mothership.
If gold does re-emerge at the core of a revitalised international monetary system and takes on a currency backing role in the future, the haphazard and non-disclosed distribution of the ECB’s current gold reserves over 5 locations, the lack of physical gold audits, and the lack of public details of any of the ECB gold holdings won’t really inspire market confidence, and is proving to be even less transparent than similar metrics from that other secretive large gold holding bloc, i.e the USA.
Welcome to the twilight zone of IMF gold sales, where transparency really means secrecy, where on-market is off-market, and where IMF gold sales documents remain indefinitely “classified” and out of public view due to the “sensitivity of the subject matter”.
Off and On Market
Between October 2009 and December 2010, the International Monetary Fund (IMF) claims to have sold a total of 403.3 tonnes of gold at market prices using a combination of ‘off-market’ sales and ‘on-market’ sales. ‘Off-market’ gold sales are gold sales to either central banks or other official sector gold holders that are executed directly between the parties, facilitated by an intermediary. For now, we will park the definition of ‘on-market’ gold sales, since as you will see below, IMF ‘on-market’ gold sales in reality are nothing like the wording used to describe them. In total, this 403.3 tonnes of gold was purportedly sold so as to boost IMF financing arrangements as well as to facilitate IMF concessional lending to the world’s poorest countries. As per its Articles of Agreement, IMF gold sales have to be executed at market prices.
Critically, the IMF claimed on numerous occasions before, during and after this 15-month sales period that its gold sales process would be ‘Transparent’. In fact, the concept of transparency was wheeled out by the IMF so often in reference to these gold sales, that it became something of a mantra. As we will see below, there was and is nothing transparent about the IMF’s gold sales process, but most importantly, the IMF blocked and continues to block access to crucial IMF board documents and papers that would provide some level of transparency about these gold sales.
Strauss-Kahn – Yes, that guy
On 18 September 2009, the IMF announced that its Executive Board had approved the sale of 403.3 metric tonnes of gold. Prior to these sales, the IMF officially claimed to hold 3217.3 tonnes of gold. Commenting on the gold sales announcement, notable party attendee and then IMF Managing Director Dominique Strauss-Kahn stated:
“These sales will be conducted in a responsible and transparent manner that avoids disruption of the gold market.”
The same IMF announcement on 18 September 2009 also stated that:
“As one of the elements of transparency, the Fund will inform markets before any on-market sales commence. In addition, the Fund will report regularly to the public on the progress with the gold sales.”
On 2 November 2009, the IMF announced the first transaction in its gold sales process, claiming that it had sold 200 tonnes of gold to the Reserve Bank of India (RBI) in what it called an ‘off-market’ transaction. This transaction was said to have been executed over 10 trading days between Monday 19 November to Friday 30 November with sales transactions priced each day at market prices prevailing on that day. On average, the 200 tonne sales transaction would amount to 20 tonnes per day over a 10 day trading period.
Note that the Reserve Bank of India revealed in 2013 that this 200 tonne gold purchase had merely been a book entry transfer, and that the purchased gold was accessible for use in a US Dollar – Gold swap, thereby suggesting that the IMF-RBI transaction was executed for gold held at the Bank of England in London, which is the only major trading center for gold-USD swaps. As a Hindu Business Line article stated in August 2013:
“According to RBI sources, the gold that India bought never came into the country as the transaction was only a book entry. The gold was purchased for $6.7 billion, in cash.”
“The Reserve Bank of India bought 200 tonnes of gold for $1,045 an ounce from the IMF four years ago. The Government can swap it for US dollars,” said [LBMA Chairman David] Gornall.”
Two weeks after the Indian purchase announcement in November 2009, another but far smaller off-market sale was announced by the IMF on 16 November 2009, this time a sale of 2 tonnes of gold to the Bank of Mauritius (the Mauritian central bank), said to have been executed on 11 November 2009. Another two weeks after this, on 25 November 2009, the IMF announced a third official sector sales transaction, this time a sale of 10 tonnes of gold to the Central Bank of Sri Lanka.
Overall, these 3 sales transactions, to the Reserve Bank of India, Bank of Mauritius and the Central Bank of Sri Lanka, totalled 212 tonnes of gold, and brought the IMF’s remaining official gold holdings down to 3005.3 tonnes at the end of 2009, leaving 191.3 tonnes of the 403.3 tonnes remaining to sell. All 3 of the above announcements by the IMF were accompanied by the following statement:
“The Fund will inform markets before any on-market sales commence, and will report regularly to the public on progress with the gold sales.”
For nearly 3 months from late November 2009, there were no other developments with the IMF’s gold sales until 17 February 2010, at which point the IMF announced that it was to begin the ‘on-market’ portion of its gold sales program. At this stage you might be wondering what the IMF’s on-market gold sales consisted of, which ‘market’ it referred to, how were the sales marketed, who the buyers were, and who executed the sales transactions. You would not be alone in wondering about these and many other related questions.
The IMF’s press releases of 17 February 2010, titled ‘IMF to Begin On-Market Sales of Gold’ was bereft of information and merely stated that the IMF would “shortly initiate the on-market phase of its gold sales program” following “the approach adopted successfully by the central banks participating in the Central Bank Gold Agreement“, and that the sales would be “conducted in a phased manner over time”. The third Central Bank Gold Agreement (CBGA) ran from September 2009 to September 2014. These CBGA’s, which have been running since September 1999, ostensibly claim to support and not disrupt the gold market but in reality have, in their entirety, been highly secretive operations where vast amounts of central bank and official sector gold is channeled via the BIS to unspecified buyers in the bullion banks or central bank space, with the operations having all the hallmarks of gold price stabilization operations, and/or official sector gold redistribution between the world’s developed and emerging market central banks.
The February 2010 announcement also made the misleading claim that “the IMF will continue to provide regular updates on progress with the gold sales through its normal reporting channels”. These regular updates have never happened.
The IMF publicly announced each official sale shortly after the transaction was concluded. A high degree of transparency will continue during the sales of gold on the market, in order to assure markets that the sales are being conducted in a responsible manner.”
However, following this February 2010 lip service to transparency, there were no direct updates from the IMF exclusively about the on-market gold sales, even after the entire gold sales program had completed in December 2010.
One further IMF ‘off-market’ gold sale transaction was announced on 9 September 2010. This was a sale of 10 tonnes of gold to Bangladesh Bank (the Bangladeshi central bank) with the transaction said to have been executed on 7 September 2010. Adding this 10 tonnes to the previous 212 tonnes of off-market sales meant that 222 tonnes of the 403.3 tonne total was sold to central banks, with the remaining 181.3 tonnes sold via ‘on-market’ transactions. The Bangladesh announcement was notable in that it also revealed that “as of end July 2010, a further 88.3 metric tons had been sold under the on-market sales announced in February 2010″. The addition of Bangladesh to the off-market buyer list that already consisted of India, Sri Lanka and Mauritius also resulted in the quite bizarre situation where the only off-market buyers of IMF comprised 4 countries that have extremely close historical, political, cultural and economic connections with each other. Three of these countries, India, Bangladesh and Sri Lanka, are represented at the IMF by the same Executive Director, who from November 2009 was Arvind Virmani, so their buying decisions were most likely coordinated through Virmani and probably through the Reserve Bank of India as well.
“The International Monetary Fund (IMF) announced today the conclusion of the limited sales program covering 403.3 metric tons of gold that was approved by the Executive Board in September 2009.”
“The gold sales were conducted under modalities to safeguard against disruption of the gold market. All gold sales were at market prices, including direct sales to official holders.”
‘Modalities’ in this context just means the attributes of the sales including the approach to the gold sales, i.e. the sales strategy. This brief announcement on 21 December 2010 was again bereft of any factual information such as which market was used for the ‘on-market’ gold sales, the identity of executing brokers, the identity of counterparties, transaction dates, settlement dates / deferred settlement dates, method of sale, information on whether bullion was actually transferred between parties, publication of weight lists, and other standard sales transaction details. Contrast this secrecy to the 1976 -1980 IMF gold sales which were conducted by a very public series auction, and which were covered in minute details by the financial publications of the time.
As usual with its treatment of official sector gold transactions, the World Gold Council’s Gold Demand Trends report, in this case its Q4 2010 report, was absolutely useless as a source of information about the IMF gold sales beyond regurgitating the press release details, and there was no discussion on how the gold was sold, who the agent was, who the buyers were etc etc.
Lip Service to Transparency
When the IMF’s ‘on-market’ sales of 191.3 tonnes of gold commenced in February – March 2010, there were attempts from various quarters to try to ascertain actual details of the sales process. Canadian investment head Eric Sprott even expressed interest in purchasing the entire 191.3 tonnes on behalf of the then newly IPO’d Sprott Physical Gold ETF. However, Sprott’s attempts to purchase the gold were refused by the IMF, and related media queries attempting to clarify the actual sales process following the IMF’s blockade of Sprott were rebuffed by the IMF.
A Business Insider article from 6 April 2010, written by Vince Veneziani and titled “Sorry Eric Sprott, There’s No Way You’re Buying Gold From The IMF”, lays out the background to this bizarre stone-walling and lack of cooperation by the IMF. Business Insider spoke to Alistair Thomson, the then external relations officer at the IMF (now Deputy Chief of Internal Communications, IMF), and asked Thomson why Sprott could not purchase the gold that was supposedly available in the ‘on-market’ sales. Thomson’s reply is summarised below:
“The IMF is only selling gold though a qualified agent. There is only one of these agents at the moment and due to the nature of the gold market, they won’t reveal who or what that agent is.”
“Sprott can’t buy the gold directly because they do not deal with institutional clients like hedge funds, pension funds, etc. The only buyers can be central bankers and sovereign nations, that sort of thing.”
The IMF board agreed months ago how they wanted to approach the sale of the gold. Sprott is welcome to buy from central banks who have bought from the IMF, but not from the IMF directly.”
While this initial response from the IMF’s Alistair Thomson contradicted the entire expectation of the global gold market which had been earlier led to believe that the ‘on-market’ gold sales were just that, sales of gold to the market, on the market, Thomson’s reply did reveal that the IMF’s ‘on-market’ gold sales appeared to be merely an exercise in using an agent, most likely the Bank for International Settlements (BIS) gold trading desk, to transfer IMF gold to a central bank or central banks that wished to remain anonymous, and not go through the publicity of the ‘off-market’ transfer process.
Although, as per usual, the servile and useless mainstream media failed to pick up on this story, the IMF’s unsatisfactory and contradictory response was deftly dissected by Chris Powell of GATA in a dispatch, also dated 6 April 2010. After discussing the IMF’s initial reply with Eric Sprott and GATA, Business Insider’s Vince Veneziani then went back to IMF spokesman Alistair Thomson with a series of reasonable and totally legitimate questions about the ‘on-market’ gold sales process.
What are the incentives for the IMF not to sell gold on the open market or to investors, be it institutional or retail?
Did gold physically change hands with the banks you have sold to so far or was the transaction basically bookkeeping stuff (the IMF still holds the physical gold in this case)?
Are there available records on the actual serial numbers of bullion? How is the gold at the IMF tracked and accounted for?
Does IMF support a need for total transparency in the sale of gold despite the effects it could have on various markets?
Shockingly, Alistair Thomson, supposedly the IMF press officer responsible for answering the public’s queries about IMF finances (including gold sales), arrogantly and ignorantly refused to answer any of the questions, replying:
“I looked through your message; we don’t have anything more for you on this.”
Another example of the world of IMF transparency, where black is white and white is black, and where press officers who have formerly worked in presstitute financial media organisations such as Thomson Reuters fit in nicely to the IMF’s culture of aloofness, status quo protection, and lack of accountability to the public.
Monthly Report on Sales of Gold on the Market
Fast forward to July 2015. While searching for documents in the IMF online archives related to these gold sales, I found 3 documents dated 2010, titled “Monthly Report on Sales of Gold on the Market“. Specifically, the 3 documents are as follows (click on links to open):
Each of these 3 documents is defined by the IMF as a Staff Memorandum (SM), which are classified as ‘Executive Board Documents’ under its disclosure policy. The IMF Executive Board consists of 24 directors in addition to the IMF Managing Director, who was in 2009 the aforementioned Dominique Strauss-Kahn. According to the IMF’s Executive Board synopsis web page, the board “carries out its work largely on the basis of papers prepared by IMF management and staff.”
The most interesting observation about these 3 documents, apart from their contents which we’ll see below, is the fact that only 3 of these documents are accessible in the IMF archives, i.e. the documents only run up to May 2010, and do not include similar documents covering the remainder of the ‘on-market’ sales period (i.e. May – December 2010). Therefore there are 7 additional monthly reports missing from the archives. That there are additional documents that have not been published was confirmed to me by IMF Archives staff – see below.
Each of the 3 reports is only 3 pages long, and each report follows a similar format. The first report spans February – March 2010, specifically from 18 February 2010 to 17 March 2010, and covers the following:
“summarizes developments in the first month of the on-market sales, covering market developments, quantities sold and average prices realized, and a comparison with widely used benchmarks, i.e., the average of London gold market fixings“
‘Market developments’ refers to a brief summary in graphical chart of the London fixing prices in US Dollars over the period in question. Quantities sold and the currency composition of sales are notable:
Sales Volume and Proceeds: A total of 515,976.638 troy ounces (16.05 metric tons) of gold was sold during the period February 18 to March 17. These sales generated proceeds of SDR 376.13 million (US$576.04 million), based on the Fund’s representative exchange rates prevailing on the day of each sale transaction.
Currency Composition of Proceeds: Sales were conducted in the four currencies included in the SDR valuation basket …., with the intention of broadly reflecting the relative quota shares of these currencies over the course of the sales program.
The 4 currencies in which the sales were conducted during the first month were USD, EUR, GBP and JPY. See table 1 in the document for more information. Perhaps the most revealing point in each document is the confirmation of the use of an agent and specifically an arrangement that the sales prices included a premium paid by the agent:
Sales Prices compared with Benchmarks: The sales were implemented as specified in the agreement with the agent. Sales were conducted at prices incorporating a premium paid by the agent over the London gold fixing, and for sales settled in currencies other than the U.S. dollar, the sales price also reflects market exchange rates at the time of the London gold fixings (10:30 am and 3:00 pm GMT), net of a cost margin.
The use of a premium over the London fixing price is very revealing because this selling strategy, where the agent paid a premium over the average London gold fixing price, is identical to the sales arrangement which the Swiss National Bank (SNB) agreed with the Bank for International Settlements (BIS) when the BIS acted as sales agent for SNB gold sales over the period May 2000 to March 2001.
As Philipp Hildebrand, ex-governor of the SNB, revealed in 2005 when discussing the SNB gold sales strategy that had been used in 2000-2001:
“At the outset, the SNB decided to use the BIS as its selling agent. Between May 2000 and March 2001, the BIS sold 220 tonnes on behalf of the SNB. For the first 120 tonnes, the SNB paid the BIS a fixed commission while the performance risk resided with the SNB. For the next 100 tonnes, the BIS agreed to pay the average price of the AM and PM London gold fixing plus a small fixed premium.“
My conclusion is therefore that the IMF also used the Bank for International Settlements in Basel, Switzerland as selling agent for its ‘on-market’ gold sales over the period February to December 2010, with the sales benchmarked to average London fixing prices in the London Gold Market.
The pertinent details for the IMF’s March – April sales document are as follows:
“A total of 516,010.977 troy ounces (16.05 metric tons) of gold was sold during the period March 18 to April 16.”
“Sales were conducted in three of the four currencies included in the SDR valuation basket” i.e. USD, EUR and JPY”
The relevant details from the April – May sales document are as follows:
“A total of 490,194.747 troy ounces (15.25 metric tons) of gold was sold during the period April 19 to May 18, 2010; no sales were conducted during the last two business days in April, owing to end of financial year audit considerations.”
“Sales were conducted in three of the four currencies included in the SDR valuation basket” i.e. USD, GBP and JPY
Purely a Pricing Exercise?
The entire ‘on-market’ gold sales program of 181.3 tonnes may well have been just a pricing exercise by the Bank for International Settlements gold trading desk to determine the market prices at which to execute the transfers, with the gold transferring ownership after the event as book entry transfers at the Bank of England in the same manner as was applied to the Indian ‘off-market’ purchase of 200 tonnes.
Taking the sales quantities in the 3 published monthly reports, and incorporating quarterly IMF gold holdings time series data from the World Gold Council, it’s possible to calculate how much gold was ‘sold’ each single day over the entire ‘on-market’ gold sales program. As it turns out, for much of the program’s duration, identical quantities of gold were sold each and every day. The ‘on-market’ program commenced on 18 February 2010. Between 18 February and 17 March, which was a period of 20 trading days in the London gold market, the agent sold 515,976.638 troy ounces (16.05 metric tons) of gold. Between 18 March and 16 April, which was also a trading period of 20 trading days (even after factoring in 2 Easter bank holidays), the agent sold a practically identical quantity of 516,010.977 troy ounces (also 16.05 metric tons). This is a daily sales rate of 25,800 ozs or 0.8025 tonnes per trading day over these 40 trading days.
During the period from 19 April to 18 May 2010, which was 19 trading days excluding the 3rd May UK bank holiday and excluding the last 2 trading days of April on which the IMF program didn’t trade, the agent sold 490,194.747 troy ounces (15.25 metric tons) of gold, which again is…wait for it… 0.8025 tonnes and 25,800 ozs per day (0.8025 * 19 = 15.2475 tonnes & 25,800 * 19 = 490,200 ozs).
Following the combined Indian, Mauritian, and Sri Lankan ‘off-market’ purchases of 212 tonnes during Q4 2009, the IMF’s gold holdings stood at 3,005.32 tonnes at the end of 2009. Based on World Gold Council (WGC) quarterly data of world official gold reserves, the IMF’s gold holdings then decreased as follows during 2010:
…resulting in total remaining gold holdings of 2,814.04 tonnes at the end of 2010, an IMF gold holdings figure which remains unchanged to this day.
These WGC figures tally with the IMF monthly report figures. For example, the IMF says that 16.05 tonnes was sold up to and including 17 March, and with another 10 trading days in March 2010, a further 8.205 tonnes (0.8025 daily sales * 10) was sold by the end of March, giving total Q1 sales of 16.05 + 8.025 = 24.075 tonnes, which is identical to the WGC quarterly change figure. The IMF was active on 59 trading days in Q2 during which it sold 47.34 tonnes, which…wait for it…was an average of 0.8024 tonnes per day (47.34 / 59 = 0.8024).
Therefore, over Q1 and Q2 2010 (i.e. between February and the end of June 2010), the ‘on-market’ sales program sold 71.42 tonnes at a consistent ~ 0.8025 tonnes daily rate. This would suggest an algorithmic program trade which offered identical quantities each and every day, or more likely just priced these quantities so as to arrive at a sales consideration amount so that the IMF would receive ‘market prices’ for its gold. Recall that IMF gold has to be sold at market prices according to the Fund’s Articles of Agreement.
Given that 88.3 tonnes had been sold ‘on-market’ by the end of July 2010 as the IMF revealed in its Bangladesh announcement, we can infer that 16.88 tonnes was sold ‘on-market’ during July 2010. This 16.88 tonne sale in July was actually at a slightly lower pace than previous months since there were 22 trading days in July 2010, however the figure was chosen due to the following: With 191.3 tonnes on sale at the outset of the ‘on-market’ program, and 71.42 tonnes sold by the end of June, this left 119.88 tonnes to sell at the end of June. Whoever was choosing the monthly sales quantities wanted to finish July with a round figure of 103 tonnes, and so chose 16.88 tonnes to sell in July (i.e. 119.88 – 16.88 = 103 tonnes). Subtracting the 10 tonnes that Bangladesh bought in September 2010 (which would have been also factored in at that time) left a round 93 tonnes (2.999 million ozs) to sell as of the beginning of August.
The Q3 2010 sales of 67.66tonnes comprised the 10 tonne ‘off-market’ sale to Bangladesh on 7 September and 57.66 tonnes of on-market sales. Given 16.88 tonnes sold in on-market sales in July, there was therefore 40.78 tonnes sold over August – September, or an average of 20.39 tonnes in each of August and September (which represented a combined 43 trading days). Overall, there were 65 trading days in Q3 and 58 trading days in Q4 (assuming that the sales wrapped up on 21 December as per the IMF announcement). From the beginning of August to the 21 December, a period of 101 trading days, the IMF sold the remaining 93 tonnes, which would be a daily sales pace of 0.93 tonnes per day.
So overall, the IMF’s 403.3 tonnes of gold sales between November 2009 and December 2010 consisted of 222 tonnes sold ‘off-market’ to India, Bangladesh, Sri lanka, and Mauritius, 88.3 tonnes sold ‘on-market’ between February and July 2010, and 93 tonnes sold ‘on-market’ between August and December 2010′.
Given that the IMF’s 4 gold depositories are the Federal Reserve Bank of New York, the Bank of England in London, the Banque de France in Paris and the Reserve Bank of India in Nagpur India, and given that the IMF gold in New York is mostly in the form of US Assay Office melts, and the gold in Nagpur is a hodgepodge of mostly low quality old gold (read non-good delivery gold), then it would be logical for the IMF to sell some of its good delivery gold which is stored in London (which, until at least the late 1970s, was predominantly held in the form of Rand Refinery 400 oz gold bars), or even in Paris, since the Banque de France has been engaged in an ongoing program of upgrading the old US Assay office gold bars in its custody to good delivery bars.
“Our bars are not all LGD [London Good Delivery quality], but we have an ongoing improvement programme.”
This Banque de France gold bar upgrading program was also confirmed in February 2011 in a National Geographic Magazine article which stated:
“Buyers don’t want the beat-up American gold. In a nearby room pallets of it are being packed up and shipped to an undisclosed location, where the bars will be melted down and recast in prettier forms.”
Top Secret Foot Notes
There are 2 interesting footnotes on page 1 or each of the 3 above documents. The first footnote states that ‘The Executive Board was briefed on the plans for on-market sales prior to the announcement’, the announcement in question being the IMF’s 17 February 2010 announcement IMF to Begin On-Market Sales of Gold.
The second footnote, which is a footnote to a sales process and sales performance summary, refers to 2 further IMF papers as follows: “Modalities for Limited Sales of Gold by the Fund (SM/09/243, 9/4/09) and DEC/14425-(09/97), 9/18/09“.
As mentioned above, SM are Staff Memorandums which are classed under Executive Board Documents. DEC series document are ‘Text of Board Decisions’ (hence the DEC) and these documents are also deemed to be Executive Board Documents. After searching for both of these documents (SM/09/243 and DEC/14425-(09/97)) in the IMF archives, it became apparent that they were not there, i.e. they were not returned and not retrievable under IMF archive search results.
This was surprisingly since the IMF claims to have what it calls its “IMF Open Archives Policy”, part of which is Article IX, Section 5, which is the “Review of the Fund’s Transparency Policy—Archives Policy“. This policy, prepared by the IMF Legal Department includes the following:
Access will be given as follows:
2. (i) Executive Board documents that are over 3 years old
(ii) Minutes of Executive Board meetings that are over 5 years old;
(iv) Other documentary materials maintained in Fund archives over 20 years old.
3. Access to Fund documents specified in paragraph 2 above that are classified as “Secret” or “Strictly Confidential” as of the date of this Decision will be granted only upon the Managing Director’s consent to their declassification. It is understood that this consent will be granted in all instances but those for which, despite the passage of time, it is determined that the material remains highly confidential or sensitive.
Given that the 2 above gold sales documents, as well as 7 other monthly reports about ‘on-market’ gold sales were missing from the archives, but all the while the IMF claimed its on-market gold sales to be “Transparent”, the next logical step was to contact the IMF Archives people and seek explanations. What follows below is the correspondence I had with the IMF Archives staff. The IMF Archives staff were very helpful and their responses were merely communicating what they had found in their systems or had been told ‘from above’. My questions and emails are in blue text. The IMF replies are in red text. My first set of queries were about the SM/09/243 and DEC/14425 documents:
02 August 2015: My first question
I’m looking for IMF document SM/09/243 “Modalities for Limited Sales of Gold by the Fund” (Sept 4th 2009) in the IMF Archives catalog (http://archivescatalog.imf.org/search.aspx). However, SM/09/243 does not appear to be in the online Archives.
But, for example SM/09/242 and SM/09/244 are both retrievable in the searchable archives, but not SM/09/243.
Can you clarify where SM/09/243 is?
02 August 2015: My second question
Could you clarify how to search for and retrieve a document in the IMF online Archives that has reference “DEC/14425-(09/97)”
This document is dated 9/18/09. I cannot find it using any of the search parameters.
3 August: IMF Archives reply
Thank you for contacting the IMF Archives. Both documents you are referring to in your recent communication, SM/09/243 and DEC/14425, are not available to the public. Please visit our website to consult on IMF Policy on Access to the Archives.
3 August: me
Can you clarify why these documents are not available to the public? i.e. have they received a certain classification?
4 August: IMF Archives
You are absolutely right, despite the time rule, these two documents are still closed because of the information security classification. We hope it answers your question.
4 August: me
Thanks for answer. Would you happen to know when (and if) these files will be available…..assuming it’s not a 20 year rule or anything like that.
5 August: IMF Archives
Could you please provide some background information about your affiliation and the need to obtain these documents. Classified documents undergo declassification process when such a request is submitted. It can be a lengthy process up to one year.
5 August: me
I was interested in these specific documents because I am researching IMF gold sales for various articles and reports that I’m planning to write.
6 Aug: IMF
Thank you for providing additional information regarding your inquiry. Please send us a formal request for the declassification of these two documents specifying your need to have access to them. We will follow through on your behalf and get back to you with a response.
Before I had replied with a formal request, the IMF archives people contacted me again on 12 August 2015 as follows:
12 Aug: IMF
While waiting for your official request we made preliminary inquiries regarding the requested documents. The decision communicated back to us is not to declassify these documents because of the sensitivity of the subject matter.
Thank you for the clarification. That’s surprising about the classification given that the IMF on-market gold sales were supposed to be transparent.
Was there any information fed back to Archives on why the ‘subject matter’ is deemed sensitive?
14 Aug: IMF Archives
“Thank you for your follow-up email. Unfortunately, these particular documents are still deemed classified and no further explanation has been communicated to the Archives.”
My next set of questions to IMF Archives in August 2015 addressed the 7 missing monthly gold sales reports that should have covered May – December 2010. Since there is a 3 year rule or maybe at max a 5 year rule under the IMF’s Transparency Policy (Archive Policy), I thought that maybe the May/June, June/July, and July/August 2010 files might be due for automatic release under the 5 year rule by the end of August 2015.
22 August 2015: Me:
“I have a question about documents which appear in the online Archive after the 5 year schedule.
Is there a scheduled update or similar which puts newly available documents in the Archive when the 5 years has elapsed?
For example, I see some documents in the Archive from June 2010, but not July/August 2010. Is there an automated process that runs, but that hasn’t yet run for July/August 2010, that puts the latest documents into the publicly available Archive?”
24 August: IMF
“Thank you for your inquiry. The review and declassification of eligible documents that meet the time rule is done by batches. Therefore, publication does not happen in real time. It is a process that takes time and might cause a delay. We will let you know when July and August documents are posted.”
2 October 2015: me
“Do you know when documents from June 2010 onwards will be added to the IMF online archive? I still don’t see any yet.
Is there a batch of declassifications for June 2010 / July 2010 / August 2010 happening soon?”
2 October: IMF
“Thank you for contacting the IMF Archives. Unfortunately, we are unable to speculate about the documents website availability and provide a more specific timeframe than the one already communicated in the attached correspondence. As already promised, we will let you know when July and August documents are posted.”
Then about 30 minutes later (on 2 October 2015) the IMF sent me another email:
2 October: IMF
“Dear Mr. Manly,
I ran a sample search of Executive Board minutes available via IMF Archives catalog and was able to find minutes issued in June and July 2010. Is there a specific document you are looking for which you are unable to find?
2 October: Me
“I was searching for the next months’ reports in the below series, report name “Monthly Report on Sales of Gold on the Market” – see screenshot attached.
The current search retrieval brings back 3 reports spanning February- May 2010, but nothing after May 2010. Report names in the retrieved search results are:
SM/10/69 SM/10/102 SM/10/139”
I was wondering if a couple of months in this series after May 2010 are available now?”
5 October: IMF
“The reports after May 2010 haven’t been declassified for public access because of the sensitivity of the subject matter, and therefore they are not available for retrieval.
We apologize for any inconvenience this may cause.”
5 October: Me
“Thanks for the reply. Out of interest, why were the reports from February to May 2010 declassified, since surely the June-December 2010 monthly reports are identical to the first three months in that they are also just providing monthly updates on the same batch of gold ~180 tonnes of gold which was being sold over the 10 month period?”
7 October: IMF
“Dear Mr. Manly,
This series of reports is under review at the moment, and according to security classification they are currently closed.
And there you have it folks. This is IMF transparency. As per the IMF Archive disclosure policy, only Christine Lagarde, current IMF Managing Director, has the authority to consent to the declassification of classified Executive Board documents.
Sensitivity of Subject Matter – China and Bullion Banks
The above IMF responses speak for themselves, but in summary, here we have an organization which claims to be transparent and which claims to have run a transparent ‘on-market’ gold sales program in 2010, but still after more than 6 years it is keeping a large number of documents about the very same gold sales classified and inaccessible to the public due to the ‘sensitivity of the subject matter’. What could be so sensitive in the contents of these documents that the IMF has to keep them classified? Matters of national security? Matters of international security? And why such extremely high level security for an asset that was recently described by the august Wall Street Journal as a ‘Pet Rock’?
The secrecy of keeping these documents classified could hardly be because of sensitivity over the way in which the sales were executed by the agent, since this was already revealed in the February – May reports that are published, and which looks like a normal enough gold sales program by the Bank for International Settlements on behalf of the IMF? Could it be to do with the identities of the counterparties, i.e. the buyer(s) of the gold? I think that is the most likely reason.
Two counterparties that spring to mind that might request anonymity in the ridiculously named ‘on-market’ sales process would be a) the Chinese State / Peoples Bank of China, and b) a group of bullion banks that were involved in gold swaps with the BIS in 2009/2010.
Chinese discretion – Market Speculation and Volatility
Bearing in mind another one of the IMF’s mantras during the 2009-2010 gold sales processes that it wanted to “avoid disruption of the gold market”, and the Chinese State’s natural surreptitiousness, the following information reported by China Daily on 24 February 2010 (which was the first week of ‘on-market’ sales) is worth considering. The article, titled ‘China unlikely to buy gold from the IMF‘, stated the following:
“Contrary to much speculation China may not buy the International Monetary Fund’s (IMF) remaining 191.3 tons of gold which is up for sale as it does not want to upset the market, a top industry official told China Daily yesterday.
“It is not feasible for China to buy the IMF bullion, as any purchase or even intent to do so would trigger market speculation and volatility,” said the official from the China Gold Association, on condition of anonymity.”
To me, these comments from the ‘anonymous’ China Gold Association official are a clear indication that if China was the buyer of the remaining 181.3 tonnes (ie. 191.3 tonnes – 10 tonnes for Bangladesh), then China certainly would have conducted the purchase in secrecy, as ‘it does not want to upset the market’, and “any purchase or even intent to do so would trigger market speculation and volatility”
In the same China Daily article, there was also a comment reported from Asian Development Bank economist Zhuang Jian, who was in favor of China buying the IMF gold, as he thought that “buying IMF gold would not only help China diversify its foreign exchange reserves but also strengthen the yuan as an international currency”, and that China would “have a bigger say in the IMF through the gold purchasing deal”.
Zhuang Jian also stated that “China can start with small purchases on the international market like the 191.3 tons of IMF gold. In the short-term, the market will see volatility, but in the long-term the prices will return to normal”.
BIS Swaps and Bullion Bank Bailouts
In late June 2010, the Bank for International Settlements (BIS) published its annual report to year-end March 2009. This report revealed that the BIS had, during its financial year, taken on gold swaps for 349 tonnes. The Wall Street Journal (WSJ) initially reported in early July 2010 that these swaps were with central banks, however the BIS clarified to the WSJ that the gold swaps were in fact with commercial banks. The Financial Times then reported in late July 2010 that “Three big banks – HSBC, Société Générale and BNP Paribas – were among more than 10 based in Europe that swapped gold with the Bank for International Settlements.” Notice that two of the named banks are French banks.
Since the BIS refuses to explain anything material about these swaps, which was most likely a gold market fire-fighting exercise, the details remain murky. But the theory that best explains what actually happened was advanced by the late Adrian Douglas of GATA in early July 2010. Douglas proposed that bullion bank gold bailout tripartite transactions actually created the BIS gold swaps. Since IMF gold is stored at both the Bank of England vaults in London and at the Banque de France vaults in Paris, IMF ‘on-market’ gold held in Paris or London would be very easy to transfer to a group of bullion banks who all hold gold accounts at the Bank of England and, it now appears, also hold gold accounts at the Banque de France.
In May 2012, George Milling-Stanley, formerly of the World Gold Council, provided some insight to the publication Central Banking about the role of the Banque de France in being able to mobilize gold. Milling-Stanley said:
“Gold stored at the Bank of England vaults … can easily be mobilised into the market via trading strategies, or posted as collateral for a currency loan”
‘Of the Banque de France, Milling-Stanley says it has ‘recently become more active in this space [mobilising gold into the market], acting primarily as an interface between the Bank for International Settlements in Basel [BIS] and commercial banks requiring dollar liquidity. These commercial banks are primarily located in Europe, especially in France’.”
It’s interesting that two of the three banks named by the Financial Times as being involved in the BIS gold swaps are French, and that Milling-Stanley mentioned that most of the commercial banks that interfaced with the BIS are French banks. Given that the then Managing Director of the IMF, Dominique Strauss-Kahn, is French, as is his successor Christine Lagarde, could some of the ‘on market’ IMF gold sales been a case of the French controlled IMF bailing out French bullion banks such as SocGen and BNP Paribas?
Applied to the IMF gold sales, and under a tripartite transaction, as I interpret it, the following transactions would occur:
IMF gold is transferred by book entry to a set of bullion banks who then transfer the title of this gold to the BIS. The BIS transfers US dollars to the bullion banks who then either transfer this currency to the IMF, or owe a cash obligation to the IMF. The sold gold is recorded in the name of the BIS but actually remains where it is custodied at the London or Paris IMF Gold Depositories, i.e. at the Bank of England or Banque de France vaults.
In this scenario, the IMF gold could have been transferred to bullion banks and further transferred to the BIS during 2009, with the ‘on-market’ pricing exercise carried out during 2010. With the BIS as gold sales agent, the entire set of transactions would be even more convenient since the BIS gold trading desk would be able to oversee the gold swaps and the gold sales.
So, in my opinion, the IMF ‘on-market’ gold on offer was either a) bought by the Chinese State, or b) was used in a gold market fire-fighting exercise to bail out a group of bullion banks, or c) a combination of the two.
Modalities of Gold Sales
As to why the IMF paper “Modalities for Limited Sales of Gold by the Fund” (Sept 4th 2009) SM/09/243″ is under lock and key and can only be declassified by the IMF Managing Director Christine Lagarde, the conclusion is that it too must contain references to something that the IMF are extremely worried about allowing into the public domain. For the simple reason is that a similarly named IMF paper from 25 June 1999, titled “Modalities for Gold Sales by the Fund” (EBS/99/110)” is accessible in the IMF Archives, and while revealing in a number of respects, it hardly contains ‘sensitive material’. This paper was prepared when the IMF had been thinking about conducting gold sales back in 1999 which never materialized, except in the form of an accounting trick to sell to and simultaneously buy back a quantity of gold to and from Mexico and Brazil. This 1999 paper “Modalities for Gold Sales by the Fund” is very interesting though for a lot of reasons as it sketches out the limitations on IMF gold sales, the approaches to the sales that were considered by the IMF at that time, and it’s also is full of pious claims that the gold sales process should be ‘transparent’, such as the following:
“it will be critical to ensure transparency and accountability of the Fund’s gold operations through clear procedures for selecting potential buyers and determining prices, and through public disclosure of the results of the sales after they have taken place. The need for transparency and evenhandedness, which is essential for an international financial institution, argues for providing as much information as possible to the public.”
On the actual approaches to gold sales, the 1999 Modalities paper introduces the topic as follows:
“This paper considers four main modalities for the sale of gold by the Fund: (i) direct sales to another official holder of gold; (ii) placements into the market through a private intermediary or a group of intermediaries, such as bullion banks; (iii) placements into the market through the intermediation of a central bank with experience in gold sales or the BIS; and (iv) direct sales to the market through public auctions, as was the case with the gold sales by the Fund between 1976 and 1980″
On the topic of publication of sales results, the 1999 paper states:
“Publication of results: In all cases, the Fund would make public at regular, say monthly, intervals the quantity sold and the prices obtained, as well as, depending on the modality decided by the Board, the names of the buyers. In the case of a forward sales strategy involving an intermediary, the Fund would make public the quantities and delivery dates of the forward sales. It would be for consideration whether the Fund would announce the names of the intermediaries selected by the Fund to sell the gold, if that modality would be chosen”
On the topic of limitations to IMF gold sales, the 1999 paper says:
“Under the Articles, the Fund is only authorized to sell gold; that is, to transfer ownership over gold on the basis of prices in the market, taking into account reasonable transactions costs. The Articles prescribe the objective of avoiding the management of the price, or the establishment of a fixed price, in the gold market (Article V, Section 12 (a)). This implies that the Fund “must seek to follow and not set a direction for prices in the gold market.“
Under the Articles, the Fund cannot engage in gold leasing or gold lending operations, enter into gold swaps, or participate in the market for gold options or other transactions that do not involve the transfer of ownership over gold.”
“Directors generally expressed the view that private placements of gold, either through a group of private institutions or through the intermediation of central banks or the BIS, had many advantages in terms of flexibility, both in terms of timing as well as in the discretion that the Fund’s agents could employ in the techniques that they could use tochannel gold into the market.“
And from the discussion, using the services of the BIS (or another central bank) appeared to be most favorable option:
“Directors further noted that there would be considerable practical difficulties in the choice of the institution or group of institutions through which the sales of gold could be conducted, even though these would be limited-but not entirely eliminated-by choosing a central bankor the BIS.“
“Greater openness and clarity by the IMF about its own policies and the advice it provides to its member countries contributes to a better understanding of the IMF’s own role and operations, building traction for the Fund’s policy advice and making it easier to hold the institution accountable. Outside scrutiny should also support the quality of surveillance and IMF-supported programs.”
“The IMF’s efforts to improve the understanding of its operations and engage more broadly with the public has been pursued along four broad lines: (i) transparency of surveillance and IMF-supported programs, (ii) transparency of its financial operations; (iii) external and internal review and evaluation; and (iv) external communications.”
“The IMF’s approach to transparency is based on the overarching principle that it will strive to disclose documents and information on a timely basis unless strong and specific reasons argue against such disclosure.”
Again, what could these “strong and specific reasons” arguing “against such disclosure” be for the 2010 IMF gold sales?
By now you will begin to see that the IMF’s interpretation of transparency on gold sales diverges massively from any generally accepted interpretation of transparency. The IMF appears to think that merely confirming that a gold sale took place or will take place is the epitome of transparency, when it would more accurately be described as obfuscation and a disdain for actual communication with the public. IMF transparency is anything but transparent.
Perhaps the usually useless mainstream financial media may finally sit up and next time they bump into the IMF’s Ms Lagarde at a press conference, ask her why the IMF continues to block access to its 2010 gold sales documents, which remain classified due to, in the IMF’s own words, “the sensitivity of the subject matter”. Here’s hoping.
Exactly 19 months to the day after the International Board of the Shanghai Gold Exchange (SGE) held its first full trading session on 19 September 2014, the Shanghai Gold Exchange launched the Shanghai Gold Benchmark Price auction on 19 April 2016. In China, the number 19 is very auspicious since it consists of lucky number 1, which means origin or beginning, and lucky number 9 meaning everlasting, eternity, or longevity.
In another example of calculated Chinese planning, the SGE first announced plans to launch its own gold fixing auction on 11 March 2015. This was the week immediately prior to the launch of the LBMA Gold Price auction on 20 March 2015, an event which occurred without any Chinese banks being present in the initial participant list. This lack of Chinese banks as initial participants in the LBMA Gold Price auctions was despite the Chinese banks having made it clear in October 2014 that they wanted to be present in the London auction on launch day:
“It’s been very welcome to see that quite a few banks in China are very interested in taking part. They said they definitely wanted to be there on day one for gold” [Ruth Crowell, LBMA CEO, October 2014 interview with MetalBulletin quoted here]
Two Chinese banks eventually joined the LBMA Gold Price auction, Bank of China on 22 June 2015, and China Construction Bank on 30 October 2015, with Industrial and Commercial Bank of China(ICBC) tee’d up to join the LBMA Gold Price auction next month on 16 May 2016. However, sources in the gold market have indicated that the Chinese banks, and others, had difficulty establishing the necessary credit lines with the incumbent bullion banks that are a LBMA perquisite for being a direct participant in the LBMA auction. This need for bilateral credit lines between auction participants is not something that the Shanghai Gold Benchmark Price suffers from, since it is using a central clearing model, something that the LBMA have paid lip-service to but that has never materialised (nor will it if the LBMA has its way).
The Shanghai Gold Benchmark Price – Details
The Shanghai Gold Benchmark Price, which I’ll abbreviate to SGE Gold Fix, is a twice daily auction held on SGE business days at 10:15 am and 2:15 pm (Beijing Time). All time zones in China are officially the same time zone (and run on Beijing Time), with Shanghai Time equivalent to Beijing Time.
The SGE Gold Fix auctions use the exchange code SHAU, and run on the electronic SGE trading platform using a ‘centralised pricing trading’ auction model. The auction is for physically-delivered 1 kg lots of 99.99% purity gold or higher, quoted in RMB per gram, with a tick size of RMB 0.01. Delivery is in the form of 1kg standard gold ingots of fineness 999.9 or higher at SGE certified vaults. For the SGE Gold Fix, standard gold is either gold from an SGE approved refinery, or gold from a LBMA approved refinery. Settlement / Delivery is two days after trade date i.e. T + 2.
At this juncture it is important to emphasise that the Shanghai Gold Benchmark Price is a centrally cleared auction on the largest physical gold exchange in the world, that delivers real physical gold bars at any of the SGE’s 55 certified vaults. Shanghai Gold Exchange uses 55 certified vaults across 36 Chinese cities for gold storage. Unlike the LBMA Gold Price auction which just settles and clears its trades as unallocated gold that merely exists as a book-keeping entry in the database tables of the LPMCL’s AURUM system.
The objective of the SGE Gold Fix auction is to arrive at a ‘Benchmark Price’, which is a price at which supply and demand reach a balance, while allowing a certain imbalance (less than 400 kgs) to remain. The overall auction concept is therefore similar to the LBMA Gold Price auctions in London. However, there are many features unique to the Shanghai auction. The SGE Gold Fix involves a ‘Reference Price’ which is used as the auction’s initial opening price. This reference price is derived from prices entered into the trading system by two specific groups of auction members during a 5 minute pre-auction window period called the ‘Reference Price Submission Window’ which runs from 10:09 am – 10:14 am for the morning auction and from 2:09 pm – 2:14 pm for the afternoon auction.
These two sets of members are ‘Fixing Members’ and ‘Reference Price Members’. All of the Fixing members are financial institutions. The Reference Price members include gold mining companies and gold jewellery companies. The logic of obtaining opening reference prices from both fixing members and reference price members is that the SGE feels it will minimise price manipulation and price collusion since the reference prices submitted include a broader set of entities (i.e. include non-financial entities). This is a clever ‘checks and balances’ approach that is lacking in the LBMA Gold Price auction.
At launch, there are 12 Fixing Members and 6 Reference Price Members. The 12 Fixing Members are all banks, 10 of which are pure Chinese banks. These 10 Chinese banks are the Big 4 state-controlled banks in the Chinese Gold Market, namely Bank of China, China Construction Bank, Industrial and Commercial Bank of China (ICBC) and Agricultural Bank of China, followed by Bank of Communications, and also Industrial Bank, Ping An Bank, Bank of Shanghai, Shanghai Pudong Development Bank, and China Minsheng Bank.
The final Fixing members are local entities of 2 foreign banks, namely Standard Chartered Bank (China) and Australia and New Zealand Bank (China) (ANZ). Both Standard Chartered and ANZ hold gold import licenses into China, as does HSBC, however, there is no indication as of yet of HSBC becoming a Fixing Member. This is despite a report in January that China would penalise in some way a foreign bank with a gold import license if it did not join the SGE Gold Fix. Two of the 12 domestic bank holding gold import licenses, Everbright and China Merchants Bank, are also absent from the SGE Gold Fix member list. Perhaps in time, they, along with HSBC will sign up.
The above is just an initial list of participants that have joined so far. The SGE maintains that any qualified entity can join up in either the Fixing of Reference Price member categories. SGE stipulates that Fixing Member applicants are required to be financially-viable financial institutions that are either active on the SGE or active in the global gold market, while Reference Price applicants can meet one of a number of criteria such as “be a leading producer or consumer in the gold industry” or “be involved in the production, processing, trading, or investment of physical gold“.
The Auction Mechanism
The Fixing members and Reference Price members submit initial reference prices. As to whether all members must submit a reference price is a moot point. Article 12.2 and 12.3 of the Rules for the Shanghai Gold Benchmark Price Trading state that the Fixing and Reference Price members “must provide market reference price at the designated time before the start of centralized pricing-trading“, however, the Shanghai Gold Benchmark Price White Paper (April 2016) describes a hierarchy of contingencies in deriving the reference price, two of which cover situations where less than 50% of members make a reference price submission.
The White Paper calculation methodology (algorithm) is as follows:
If > 50% of members submit a reference price, SGE calculates an arithmetic mean after disregarding the highest and lowest submitted price.
If < 50% of members submit a reference price, the SGE calculates an average (arithmetic mean) of all trades in the Au9999 spot gold contract that have been executed on the SGE during the timeframe for submitting reference prices.
If no trades were executed in the AU9999 during that time, the SGE takes the Shanghai Gold Reference Price from the previous trading session as the initial price. [this would be the previous afternoon benchmark price if applied to the morning pre-auction etc]
The Au9999 is the SGE busiest spot gold contract. Based on this three-pronged approach, it would seem that the members are not all obliged to submit a reference price, otherwise the 50% threshold would never arise unless due to communication outages or similar. The only logical interpretation of the two documents is that if a member turns up to the auction (or logs in to the trading platform), then they are obliged to submit a reference price. If they don’t turn up, then there is no obligation. Notwithstanding this grey area, after the reference price is calculated the SGE then publishes the opening price.
Some readers will recall that ICE Benchmark Administration (IBA) uses a ‘human’ chairperson to come up with the opening price in the LBMA Gold Price auction using a number of price sources that ICE Benchmark Administration will not divulge. Nor will ICE Benchmark Administration divulge the identities of the panel of chairpersons that it employs to chair the daily LBMA Gold Price auctions. Frankly, this is a disgrace and a scandal, and shows that the Chinese auction methodology is far more transparent that its London counterpart. My hunch is that there are names involved as chairpersons in the current LBMA Gold Price auction that were also involved in the former London Gold Market Fixing Limited company which operated the London Gold Fixing auctions. Otherwise, why keep the identities a secret. No mainstream financial journalists in London will touch this particular story, although they are all aware of it. See BullionStar blog “Six months on ICE – The LBMA Gold Price” for further details about the lack of transparency in the administration of the LBMA Gold Price auction.
Once the opening price of the Shanghai Gold Benchmark Price is established using the calculated reference price, the auction begins, and participants and their clients submit their buy or sell orders and transaction volumes etc. The auction consists of a first round and possible subsequent rounds if supply and demand don’t reach a balance. There are two distinct time periods in each round, a ‘market tendering‘ session and a ‘supplementary tendering‘ session. The market tendering part is just the normal part of the round where all participants and their clients submit orders. The supplementary tendering session in each round only applies to the Fixing members, and allows them to submit supplementary orders against the remaining imbalanced quantity so as to try to reduce the imbalance to less than 400 kgs and so speed up the auction, because if the imbalance is shrunk to under 400 kgs, there is no need for an additional round(s).
The first round consists of a 1 minute market tendering session + a supplementary tendering session of 10 seconds. If the price is not balanced after the first round, the SGE trading system will adjust the price upwards or downwards depending on buy and sell orders, and then a new round begins. Any and all subsequent rounds consist of 30 seconds duration of a market tendering session + 10 seconds of a supplementary tendering session.
Once the imbalance is less than 400 kgs, it is shared out among the Fixing members. The price is then said to be balanced and the SGE then publishes the benchmark price. The ‘Shanghai Gold Benchmark Price’ now has its own web page on the SGE website here, with daily price lookup, daily, monthly and annual charting (which will make sense when the auction has been running for a while), and Trading Rules, Contract Spec and Q & A (to be uploaded, but some of which are already detailed in the Rules and White Paper linked above).
SGE Surveillance Committee
The SGE has also created an Oversight Committee to monitor and oversee the auction’s functioning. This Committee currently comprises 11 representatives from 9 organisations. Although the names of the representatives have not yet been published, the names of their organisations have. The SGE will have 3 representatives, and the 8 other entities will each have 1 representative. The list is as follows
Bank of China 1
Standard Chartered Bank (China) 1
ANZ Bank (China) 1
China Gold Coin Corporation 1
Baird Mint 1
China Gold Association 1
World Gold Council 1
Of the list of 11, seven reps come from pure Chinese entities, with the remaining four from two foreign banks, the World Gold Council and ‘Baird Mint’. All represented entities have connections with the SGE except it seems Baird.
The Oversight Committee’s remit is to monitor trading, clearing, delivery, in terms with SGE rules, analyse trading behaviour, examine conflicts of interest etc.
The SGE uses central clearing of the trades executed in the SGE Gold Fix auctions and so there is no credit risk between participants. Under central clearing, the exchange becomes the counterparty to all buyers and sellers. This also avoids the need for participants to maintain bilateral credit arrangements with each other, and so easily allows the number of auction participants to grow, even exponentially. The lack of central clearing in the LBMA Gold Price auction is a huge barrier to entry for non-bullion bank participants and has been kept as such by the LBMA, even though ICE offers central clearing and has been well able to implement a centrally cleared model from Day 1 in March 2015. See ICE Executive Summary which summarises the winning ICE bid for the LBMA Gold Price wherein ICE discusses “moving to a centrally cleared model“.
The Purpose of the Shanghai Gold Benchmark Price
The Shanghai Gold Benchmark Price is but one more step in the growth and deregulation of the Chinese gold industry, and the internationalisation and extended use of the RMB. Much of this scene was set back in 2000 at the China Gold Economic Forum. It’s also a natural step for a country that is the largest gold producer, gold consumer and gold importer on earth. The SGE Gold Fix also provides China with a real role in global gold price discovery and creates the first proper transparent RMB denominated gold price benchmark, calculated within a centralised trading auction setting on an exchange.
An RMB gold price benchmark aids risk management and hedging in the domestic gold sector, and can also now be used within Chinese gold-backed derivative products, a function which the SGE has explicitly mentioned. So expect financial products to appear that use the Shanghai Gold Benchmark Price as a reference or valuation price. In China, where gold is correctly recognised as the ultimate money, there is also the prestige of having an internationally known global gold price benchmark, that will, in SGE’s words “enhance China’s voice in the global gold pricing market”.
In its White Paper, the SGE states that “the relationship between Shanghai Gold and Loco London Gold is non-competitive”, and it lists a number of reasons why this, on paper, is so, such as the London auction is for the OTC trading of 400 oz bars of 99.5 purity quoted in USD, while the Shanghai auction is Exchange-based trading for 1 kg bars of 99.99 purity quoted in RMB. While this is true, these are only ‘contract spec’ differences, and having a PBoC controlled gold benchmark that is not in London and not under the control of LPMCL clearing banks and the Bank of England is a much bigger change than purely differing contract specs.
The Chinese play a long patient game and more often than not just go ahead and do things / make things instead of just talking about doing things. The SGE and the PBoC have now set up another part of the infrastructure that can in time play a critical role in the global gold market as the Renminbi begins to internationalise. Whenever the Chinese Government and PBoC move to allow gold to be officially exported, this will really boost the new kid on the block benchmark.
I would not think that the Chinese will want to make waves with this Shanghai benchmark in the near future that would explicitly jeopardise their relationships with the London Gold Market. The fact that the luminaries of the global gold world were at the SGE Gold Fix launch ceremony and the China Gold Market Summit Forum on 19 April, much like they were at the launch of the SGE International Board in September 2014, attests to the fact that large players such as the World Gold Council, LBMA, MKS, ANZ and Standard Chartered are very much in a cooperative relationship with the SGE, the China Gold Association and the large Chinese banks. As the Chinese Gold Market continues to evolve, my view is that the Shanghai Gold Price Benchmark will naturally move into the ascendancy, and that its physical gold price discovery influence will subtly begin to show up the London Gold Market’s trading weaknesses (i.e. small % of physical traded), or alternatively, the Chinese will at some stage call time-out when ready, and allow the Shanghai Gold Price Benchmark to really shift up a gear to generate physical gold prices that will disconnect from the COMEX and LBMA pass the parcel shenanigans.
That article highlighted that the amount of gold stored in custody at the Bank of England (BoE) fell by 350 tonnes during the year to 28 February 2015, after also falling by 755 tonnes during the year to end of February 2014. Therefore, by 28 February 2015, there was, according to the BoE’s own statement, £140 billion or 5134.37 tonnes of gold in custody of the BoE, or in other words ~ 410,720 Good Delivery gold bars.
The article also reviewed snapshots of the total amount of gold stored in the London vaults at various recent points in time.
Firstly, a reference on the London Bullion Market Association (LBMA) web site for a date sometime before 2013 stated that there had been 9,000 tonnes of gold (i.e. 720,000 Good Delivery bars) stored in London with two-thirds of this amount, or 6,000 tonnes, stored in the Bank of England (about 482,000 bars), and 3,000 tonnes stored in London ex Bank of England vaults (238,000 bars). (Nick Laird of Sharelynx subsequently pointed out to me that the earliest reference to this 9,000 tonne figure was from a LBMA presentation from November 2011.)
Secondly, by early 2014, the LBMA web site stated that there were only 7,500 tonnes of gold in all London vaults, i.e. ~600,000 bars, and of this total, three-quarters or 5,625 tonnes were at in Bank of England, ~ 450,000 bars, and only one-quarter or 1,875 tonnes was stored at LBMA London gold vaults excluding the Bank of England’s gold vaults.
So, the entire London market including the Bank of England had lost 1,500 tonnes (120,000 bars) between 2011 and early 2014, with 375 tonnes less in the BoE and 1,125 tonnes less in the London market outside the BoE.
Finally, on 15 June 2015, the LBMA stated that “There are ~500,000 bars in the London vaults, worth a total of ~US$237 billion”. This ~ 500,000 bars equates to 6,256 tonnes. (On 15th June 2015, the morning LBMA Gold Price was set at $1178.25, which would make $237 billion worth of gold equal to 201.145 million ounces, which is 6,256 tonnes).
Therefore, another ~1,250 tonnes of gold (approximately 100,000 Good Delivery bars) departed from the London gold vaults compared to the early 2014 quotation of 7,500 tonnes of gold in the London vaults.
So overall, between the 9,000 tonnes quotation in 2011, and the 6,256 tonnes 2015 quotation, some 2,750 tonnes (~ 220,000 Good Delivery bars) disappeared from the London gold vaults. With 6,256 tonnes of gold stored in the entire London vault network in 2015, and with 5,134 tonnes of this at the Bank of England, that would leave 1,122 tonnes of gold in London outside the Bank of England vaults.
To reiterate, “the London gold vaults“, in addition to the Bank of England gold vaults, refer to the storage vaults of JP Morgan and HSBC Bank in the City of London, the vaults of Brinks, Malca Amit and Via Mat (Loomis) located near London Heathrow Airport, the vault of G4S in Park Royal, and the Barclays vault managed by Brinks.
Because the Bank of England reveals in its annual report each year the value of gold it has stored in custody for its customers (central banks, international official sector institutions, and LBMA member banks), then it is possible to compare 3 years of gold tonnage figures, namely the years 2011, 2014 and 2015, and then show within each year how much of this gold is stored at the Bank of England, and how much is stored in London but outside the Bank of England vaults.
Nick Laird of www.sharelynx.com / www.goldchartsrus.com has done exactly this in the following sets of fantastic charts which he has created to graphically capture the above London gold trends, and a lot more besides. These charts are just a subset of a suite of inter-related gold charts that Nick has created to address this critical subject in the London Gold Market.
Although the Bank of England is not a LBMA member, the Bank of England gold vaults are a critical part of the LBMA gold vaulting and gold clearing system, and LBMA bullion banks maintain gold accounts with the Bank of England which facilitate, among other things, gold lending and gold swaps transactions with central banks. Hence the above and below charts are titled “LBMA Vaulted Gold in London”.
My “How many Good Delivery gold bars are in all the London Vaults” article had also quantified that nearly all of this ~1,122 tonnes consists of gold from physical gold-backed ETFs which store their gold in the London vaults. (previously rounded up to 1,125 tonnes for ease of calculation).
I had included 5 gold ETFs in my previous analysis namely, SPDR Gold Trust (GLD), Shares Gold Trust (IAU), ETF Securities – ETFS Physical Gold ETF (PHAU & PHGP), ETF Securities – Gold Bullion Securities (GBS & GBSS), and Source Physical Gold ETC (P-ETC), and also some smaller holdings at BullionVault and GoldMoney. In total these ETFs and other holdings accounted for just over 1,000 tonnes of gold in the London market.
However, I had missed a few other gold ETFs which also store their gold in the London vaults. Nick Laird, whose Sharelynx website maintains up-to-date gold ETF data and gold holdings, took the initiative to fill in the missing ETF blanks and Nick re-calculated the more comprehensive ETF holdings figures for London, which worked out at an exact 1,116 tonnes of gold, astonishingly close to the implied figure represented by the 1,122 tonnes outside the Bank of England vaults.
The additional gold backed ETFs also included in Nick Laird’s wider catchment were Deutsche Bank db Physical Gold ETC and associated Deutsche ETFs, ABSA gold ETF (of South Africa), Merk Gold ETF, and some smaller holdings from Betashares and Standard Bank. The following chart from Sharelynx shows the full data for physically backed gold ETFs storing their gold in London:
We then discussed an approach, in conjunction with Koos Jansen and Bron Suchecki, to identify known central bank gold stored in the Bank of England vaults by tallying up this storage data on a country level basis. So, for example, assuming 5,134 tonnes of gold stored at the Bank of England in early 2015, the aim would be to try to account for as much of this gold as possible using central bank sources.
As mentioned in the ‘How many gold bars‘ article, the Bank of England stated in 2014 that 72 central banks (including a few official sector financial organisations) held gold accounts with the Bank. It is not known if any of these gold accounts are inactive or whether any of these accounts have zero gold holdings. The LBMA stated in 2011 that “The Bank of England acts as gold custodian for about 100 customers, including central banks and international financial institutions, LBMA members and the UK government”. Therefore there could also be more than 25 LBMA member commercial banks with gold accounts at the Bank of England.
Some of the Bank of England 5,134 tonne total would therefore be gold held in LBMA member bank gold accounts at the Bank of England, for which data is not public. Likewise, a lot of central banks do not reveal where their gold is stored, let alone how much is stored in specific vaults such as at the Bank of England and Federal Reserve Bank of New York.
However, many central banks have more recently begun to provide some information on where they say their official reserve gold is stored. Other central banks have always been to some extent transparent. Overall, a variety of sources, where possible, can be used to source locational data regarding central bank gold storage locations. There will continue to be gaps however, since some central banks remain non-cooperative, even when asked directly about where they stored their gold.
Tallying this type of central bank gold storage data will probably be a work in progress. However, there has to be a cut-off point for doing a first pass through the data, and this is a first pass. As a group, the European central banks have been especially forthcoming with gold storage data, compared to even 3-4 years ago (except for Spain). For other central banks, I looked in various places such as their financial accounts, and I contacted some of them by email with varying degrees of success. About half of the 72 central banks on the Bank of England’s list were identified, again, with varying degrees of accuracy.
The following fantastic chart by Nick Laird captures an overview of this Bank of England gold storage data. Essentially the chart shows that the banks listed hold, or have stated that they hold, the respective quantity listed, and in total the named banks could account for x tonnes gold stored at the bank of England. This is labelled ‘Known Gold‘. Given ‘Known Gold’, this leaves the residual as ‘Unknown Gold‘.
The remainder of this article explains the logic and the sources behind each country, and why that country appears on the list. When a central bank claims to have stored gold at the Bank of England, or the evidence suggests that, it does not necessarily mean that the gold in question is held in custody in a gold set aside account or that it is allocated in identifiable bars, or even that it is actually there. Many central banks engage in gold lending, or have done so in the last 15-20 years, and have at times, or permanently, transferred control of that gold to LBMA bullion banks.
Until all central banks come clean about what form their gold holdings are in, which will never happen, then the amount of central bank gold that’s encumbered by bullion banks or under claims, liens, loan agreements etc will not be apparent.
Germany holds 3,384 tonnes of gold, and 12.9%, or 438 tonnes are stored at the Bank of England. The Bundesbank’s ongoing repatriation of gold from New York and Paris does not alter the amount of Bundesbank gold held at the Bank of England.
“Most of Danmarks Nationalbank’s gold is stored at the Bank of England, where it has been since it was moved for safety reasons during the Cold War. In March 2014, Danmarks Nationalbank inspected its stock of gold in the Bank of England.”
Therefore, the assumption here is that 62.7 tonnes of Danish gold is stored at the Bank of England.
Note the Danmarks Nationalbank’s assertion that in order for gold to be lent it has to be moved to the London, since London is the centre of the gold lending market.
In 1999 “Almost 99 per cent, or 93 per cent of the Nationalbank’s total gold stock, had been lent.” The same 1999 Danish central bank article also said that:
I have underlined the above sentence since it’s of critical importance to understanding that in gold lending, central bank gold lent to LBMA bullion banks at the Bank of England does not necessarily move out of the Bank of England vaults. Lent gold may or may not move out the door, depending on what the borrower plans to do with the borrowed gold.
It also means that the total gold in custody figure that the Bank of England reveals each year (for example £140 billion in February 2015), consists of:
a) central bank gold stored at the Bank of England
b) bullion bank gold stored at the Bank of England
c) central bank gold that has been lent or swapped with bullion banks (gold deposits and gold swaps) and that has not been moved out of the Bank of England vaults. This category of gold is still in custody at the Bank of England. The central bank claims to still own it, the bullion bank has control over it, and the Bank of England still counts it as being in its custody.
The Netherlands holds 612.5 tonnes of gold, and 18%, or 110 tonnes are stored at the Bank of England.
Notice that the UK gold reserves includes holdings of gold coin, as well as gold bars.
Ireland hold 6 tonnes of gold in its official reserves, a small amount of which is in the form of gold coins, but nearly all of which is in the form of gold bars stored at the Bank of England.
Recently, I submitted a Freedom of information (FOI) request to the Central Bank of Ireland requesting information such as a weight list of Ireland’s gold stored at the Bank of England. After the FOI request was refused and the Central Bank of Ireland claimed there was no weight list, I appealed the refusal and was provided with a SWIFT ‘account statement’ from 2010 that the Bank of England had provided to the Central Bank of Ireland. See below:
This statement shows that as of 31 December 2010, the Central Bank of Ireland held 453 gold bars at the Bank of England with a total fine ounce content of 182,555.914 ounces, which equates to an average gold content of 402.993 fine ounces per bar. It also equates to 5.678 tonnes, which rounded up is 5.7 tonnes of gold stored at the Bank of England.
The fact that no weight list could be tracked down is highly suspicious, as is the fact that Ireland had in earlier years engaged in gold lending, so did not, at various times in the 2000s have all of its gold allocated in the Bank of England. How a central bank can claim to hold gold bars but at the same time cannot request a weight list of those same bars is illogical and suggests there is a lot more that the Central Bank of Ireland will not reveal.
Belgium holds 227 tonnes of gold, most of which is stored at the Bank of England with smaller amounts held with the Bank of Canada and with the Bank for International Settlements. Banque Nationale de Belgique (aka Nationale Bank van België (NBB)) does not publish an exact breakdown of the percentage stored at each location, however, in March 2013 in the Belgian Parliament, the deputy Prime Minister and Minister for Finance gave the following response in answer to a question about the Belgian gold reserves:
“Most of the gold reserves of the National Bank of Belgium (NBB) is indeed held with the Bank of England. A much smaller amount held with the Bank of Canada and the Bank for International Settlements. A very limited amount stored in the National Bank of Belgium.”
Furthermore, there were a series of reports in late 2014 and early 2015 that would suggest that Belgium stores 200 tonnes of its gold at the Bank of England. Firstly, in December 2014, VTM-nieuws in Belgium reported that the NBB governor Luc Coene had said that the NBB was investigating repatriating all of its gold. See Koos Jansen article here.
On 4 February 2015, Belgian newspaper Het Nieuwsblad said that Belgium would repatriate 200 tonnes of gold from the Bank of England, but the next day on 5 February 2015, another Belgian newspaper De Tijd reported that NBB Luc Coene denied the repatriation report, and quoted him as saying:
“There are other and more effective ways to verify if the gold in London is really ours. We have an audit committee that inspects the Belgian gold in the UK regularly”.
Therefore, the assumption here, backed up by evidence, is that Belgium stores 200 tonnes of gold at the Bank of England.
Australia holds approximately 80 tonnes of gold in its official reserves, with 1 tonne on loan, and 99.9% of gold holdings stored at the Bank of England. See 2014 annual report, page 33. According to a weight list of its gold held at the Bank of England, released via an FOIA request in 2014, Australia stores approximately 78.8 tonnes of gold at the Bank of England.
South Korea (Bank of Korea) holds 104.4 tonnes of gold, 100% of which, or 104.4 tonnes is stored at the Bank of England. The Bank confirmed this to me in an emailon 11 September 2015. See email here ->
International Monetary Fund
The IMF currently claims to hold 2,814 tonnes of gold after apparently selling 403.3 tonnes over 2009 and 2010 (222 tonnes in ‘off-market transactions and 181.3 tonnes in ‘on-market transactions’). Prior to 2009, IMF gold holdings had been 3,217 tonnes, and had been essentially static at this figure since 1980 [In 1999 IMF undertook some accounting related gold sale transactions which where merely sale and buyback bookkeeping transactions].
Although the IMF no longer provide a breakdown of how much of its gold is stored in each location where it stores gold, the amount of gold held by the IMF at the Bank of England can be calculated by retracing IMF transactions from a time when the IMF did provide such details. In January 1976, the IMF held 898 tonnes of gold at the Bank of England in London, 3,341 tonnes at the Federal Reserve Bank of New York, 389 tonnes at the Banque de France in Paris, and 144 tonnes at the Reserve Bank of India in Nagpur, India. Therefore, of the IMF’s total 4,772 tonnes holdings at that time, 70% was stored in New York, 19% in London, 8% in Paris and 3% in India. See here and here.
In the late 1970s, the IMF sold 50 million ounces of gold via two methods, namely, 25 million ounces by ‘public’ auctions, and 25 million ounces by distributions to member countries.
In the four-year period between mid-1976 and mid-1980, the IMF sold 25 million ounces of gold to the commercial sector via 45 auctions. Thirty five of these auctions delivered gold at the FRBNY, 7 of these auctions delivered gold at the Bank of England, and 3 of the auctions delivered gold at the Banque de France.
Of the 7 auctions that delivered the IMF’s gold at the Bank of England, these auctions in total delivered 3.74 million ounces [Dec-76: 780,000 ozs, Aug-77: 525,000 ozs, Nov-77: 525,000 ozs, May-78: 525,000 ozs, Oct-78: 470,000 ozs, Mar-79: 470,000, and Dec 79:444,000 ozs], which is 116 tonnes. See IMF annual report 1980.
The IMF also sold 25 million ozs of gold to its member countries within four tranches over the 3 year period from January 1977 to early 1980. These sales, which were also called gold ‘distributions’ or ‘restitutions’ and covered between 112 and 127 member countries across the tranches, were initially quite complicated in the way they were structured since they involved IMF rules around quotas which necessitated the gold being transferred to creditor countries of the IMF and then transferred to the purchasing countries. In the later sales in 1979 and 1980 countries could purchase directly from the IMF.
Countries could choose where to receive their purchased gold, i.e. London, New York, Paris or Nagpur, however, the US, UK, France and India, which had the largest IMF quotas and hence the largest gold distributions, all had to receive their gold at the respective IMF depository in their own country. I don’t have the distribution figures to hand at the moment for the 25 million ozs sold to countries, but about 18 countries took delivery from the Banque de France in Paris, with the rest choosing delivery from New York and London.
Therefore an assumption is needed on the amount of gold the IMF ‘distributed’ to member countries from its Bank of England holdings between 1977 and 1980. Of the 25 million ounces distributed, the US received 5.734 million ozs, the UK received 2.396 million ozs (75 tonnes), France received 1.284 million ozs, and India received 805,000 ozs. Subtracting all of these from 25 million ozs leaves 14.78 million ozs which was distributed to the other ~120 countries. Since the IMF held 70% of its holdings at the FRBNY in 1976, 19% at the Bank of England and 8% at the Banque de France, apportioning these three weights to the remaining 14.78 million ozs would result in 10.76 million ozs (332 tonnes) being sold from the FRBNY, 2.867 million ozs (89 tonnes) from the Bank of England and 1.24 million ozs (38.5 tonnes) from the Banque de France.
Adding this 89 tonnes to the 75 tonnes received by the UK would be 164 tonnes distributed from the Bank of England IMF gold holdings. Add to this the 116 tonnes of London stored IMF gold sold in the auctions equals 280 tonnes. Subtracting this 280 tonnes from the IMF’s London holdings of 898 tonnes in January 1976 leaves 618 tonnes.
In 2009 the IMF said that it had sold 200 tonnes of gold to India, 2 tonnes to Mauritius, 10 tonnes to Sri Lanka,and then 10 tonnes to Bangladesh in 2010. The Bangladesh figures reflect its 10 tonne purchase. However, at the moment, there has been no exact confirmation that the 200 tonnes that India bought is in London. It probably is in London, but leaving this amount under the IMF holdings instead of in India’s holdings makes no difference. Subtracting the Bangladesh sale of 10 tonnes, and rounding down slightly, there are 600 tonnes of IMF gold (excluding the 2009 India 200 tonnes sale) storedat the Bank of England.
The IMF sales of gold to Sri Lanka and Mauritius in 2009 of a combined total of 12 tonnes probably came out of the IMF’s London holdings also. The IMF’s sale of 181.3 tonnes of gold in 2010 via ‘on-market transactions’ may also have come out of the IMF’s London stored gold. These ‘on-market transactions” look to have used the BIS as pricing agent, and the IMF have gone to great lengths to hide the full details of these sales from public view. More about that in a future article.
The Reserve Bank of India holds 557.75 tonnes of gold. Of this total, a combined 265.49 tonnes are stored (outside India) at the Bank of England and with the Bank for International Settlements. In 2009 India purchased 200 tonnes of gold from the IMF via an ‘off-market transaction‘. A slide from this presentation sums up this information.
The questions then are, is the 200 tonne purchase from the IMF stored at the Bank of England, and how much of the earlier 65.49 tonnes is stored at the Bank of England.
A 2013 article in the Indian Business Standard which was reprinted from “Reserve Bank of India history series. Volume 4, 1981-1997, Part A”, explains that in 1991, the Reserve Bank of India entered 2 separate gold loan deals, one deal with UBS in Switzerland (which required 18.36 tonnes of RBI gold to be sent to Switzerland) and the other deal with the Bank of England and Bank of Japan (where 46.91 tonnes was required to be sent to the Bank of England). Together those 2 transactions equals 65.27 tonnes which is 0.222 tonnes short of the 65.49 total.
After the gold loan deals expired, it looks like 18.36 tonnes of Indian gold were left in Switzerland and transferred to safekeeping or deposit with the BIS, and 46.91 tonnes of Indian gold was left at the Bank of England.
Regarding India’s purchase of 200 tonnes of gold in 2009, the IMF only has gold 4 depositories, namely, the Bank of England, Federal Reserve Bank of New York, Banque de France, and the Reserve Bank of India in Nagpur, India. Given that the Indian gold stored abroad is “with the Bank of England and the Bank for International Settlements“, then for the 200 tonnes of IMF gold to end up being classified as ‘with’ the BIS, it would have to have either been transferred internally at one of the IMF depositories to aBIS account, or transferred via a location swap or a physical shipment to a BIS gold account at the vaults of the Swiss National Bank in Berne.
For now, the 200 tonnes of gold sold by the IMF to India in 2009 is reflected in the IMF holdings and not the India holdings. It does not make a difference to the calculations, since the 200 tonnes is still at the Bank of England.
Bulgaria has 40.1 tonnes of official gold reserves. The latest BNB annual report states that 513,000 ozs are in standard gold form, and 775,000 ozs are in gold deposits.
The Bank for International Settlements (BIS), headquartered in Basle, Switzerland does not have run any gold vaults of its own. However, the BIS is a big player in the global central bank gold market, and it offers its central bank clientele gold safekeeping (and settlement) services using central bank vaults in London, New York and Berne. These services are possible because the BIS maintains gold accounts at the Bank of England, the Federal Reserve Bank of New York, and the Swiss National Bank in Berne. BIS gold accounts can act like omnibus accounts in that many central banks can hold gold in sub-accounts under a BIS gold account at each of these institutions in London, New York and Berne.
Gold can then be transferred around locations using gold swaps where one of the counterparties to the gold swap is the BIS.
The BIS is involved with gold in 3 main categories.
a) the BIS holds gold in custody for customers, off of the BIS balance sheet
b) the BIS has its own gold holdings which are classified as its gold investment portfolio, and which are on its balance sheet
c) the BIS accepts gold deposits from central banks. These gold deposits appear as a liability on the BIS balance sheet. Then the BIS turns around and places these gold liabilities in the market under its own name. These placing are also in the form of gold deposits and gold loans with other institutions including commercial banks. These ‘assets’ are then classified on the BIS balance sheet as BIS’ “gold banking” assets.
a) In its latest annual report, as of the end of March 2015, the BIS stated that it holds 443 tonnes of gold under earmark for its central bank customers on a custody basis. This gold is not on the BIS balance sheet. i.e. it is ‘off-balance sheet’ gold held by the BIS.
b) The BIS also holds 108 tonnes of its own gold (on balance sheet within an investment portfolio). This BIS gold is either kept in custody or transferred to bullion banks as gold deposits. The BIS does not provide granular data in its annual report as to how much of its own gold is ever put into gold deposits.
c) As of 31 March 2015, the BIS had 510 tonnes of gold assets on its balance sheet. Of this total, 108 tonnes was the BIS’ own gold, leaving 403 tonnes as banking assets (i.e. customer gold . Of this same 510 tonnes total, 55 tonnes were classified as gold loans, so 457 tonnes were not gold loans. If all 55 tonnes of gold loans were from customer gold, this would leave 348 tonnes of customer backed gold banking assets. On the same date (31 March 2015), the BIS held 356 tonnes of gold deposits from customers (sight deposits and short-term deposits) on the liability side of its balance sheet which originate entirely from central banks depositing gold with the BIS in sight and term deposits.
The question then is how to reflect BIS gold storage holdings at the Bank of England. While most if not all gold deposit transactions between central banks/BIS and bullion banks take place in London, the data is not readily published.
It was therefore decided, in the spirit of being conservative, to make an assumption on the BIS gold, and only use BIS customer custody gold and BIS own gold as inputs, and because BIS has gold accounts with 3 vaults (London, NY and Berne), to then just divide by 3 and say that one-third of BIS own gold and one-third of BIS ‘central bank custody gold’ is in London This would be 183.66 tonnes, i.e. (108+443)/3.
Therefore, this model states that 183.66 tonnes of BIS gold is stored in the Bank of England. This is probably being very conservative, especially given that no on-balance gold deposited by BIS customers is reflected in this figure.
In September 2010, the IMF sold 10 tonnes of gold to Bangladesh Bank, bringing total gold holdings up from 3.5 tonnes to 13.5 tonnes. The fact that this gold is stored at the Bank of England shows that the IMF sold this gold from its holdings that were stored at the Bank of England. (Note, Bangladesh has recently added some small amounts of domestic confiscated gold to its reserves).
Mexico’s central bank, Banco de Mexico (Banxico) currently hold 122.1 tonnes of gold. At the end of 2012, Mexican official gold reserves totalled 4,034,802 ounces (125 tonnes), of which only 194,539 ounces (6 tonnes) was in Mexico, and 119 tonnes abroad.
With Banxico now holding 122 tonnes according to the World Gold Council, and not 125 tonnes, the assumption is that the 3 tonne reduction came from domestic holdings.
Poland holds 102.9 tonnes of gold in its reserves. Poland’s central bank (Narodowi Bank Polski (NBP)) published a guide to Poland’s gold in 2014 in which it confirmed that nearly all of its gold is at the Bank of England. See pages 86-90 of the guide.
“How much gold did Poland possess before 1998? Approximately 746,463 ounces, of which almost 721 thousand was invested in deposits in commercial banks. In turn, the gold kept in the country was mainly coins, gold bars and various types of gold “scrap” bought by NBP.” (page 86)
Before 1998, only 25,463 ozs of NBP gold was kept in Poland, and 721,000 ozs (22.43 tonnes) was deposited with bullion banks. Poland then bought 80 tonnes of gold in 1998, bringing its gold reserves up to nearly 103 tonnes. The purchase was done as follows:
“…we used the services of a bank which constantly carries out similar transactions. Next, we made a location swap and the whole of NBP’s foreign gold reserves were deposited onto our account in the Bank of England.” (page 88)
It is likely that the NBP is referring to the BIS as the bank which purchased the gold on behalf of Poland, and then transferred it from one of the BIS gold accounts at the Bank of England to the NBP gold account at the Bank of England.
So that is 102.9 tonnes stored at the Bank of England.
Note also that, the Polish central bank explains that “It can be assumed that the gold that has been placed on the market at any time is precisely the gold that is held by the central banks in London“. In other words, central banks that have places gold on deposit (lent it) have done so with gold that they have stored in the Bank of England. See the following screenshot:
Note 6.1 on page 136 of the 2013 NBP annual report states:
“Gold and gold receivables The item comprises gold stored at NBP and deposited in a foreign bank account. As at 31 December 2013, NBP held 3,308.9 thousand ounces of gold (102.9 tonnes).”
This statement about the “gold stored at NBP and deposited in a foreign bank account” has been in a few of the recent NBP annual reports. In April 2013, before the NBP had published the guide to its gold, I asked the NBP by email, based on the statement, to clarify if the gold held abroad is held in custody, for example at the Bank of England or FRBNY or held in time deposits with commercial banks?”
The NBP responded: “Narodowy Bank Polski does not make gold time deposits with commercial banks”.
This may be true if the NBP is using sight deposits, but the 2013 answer, like so many other central banks currently, avoided providing any real information to the question.
Given that nearly all NBP’s 102.9 tonnes of gold was in the Bank of England when the 80 tonnes purchase was made in 1998, the assumption here is that still is the case, and that for simplicity, 100 tonnes of Poland’s gold is at the Bank of England.
Romania has 103.7 tonnes of gold in its official reserves.
In percentage terms, as at 31 December 2014, 27% of Romania’s gold was in ‘standard form’ which presumably means Good Delivery Bars (400 oz bars), 14% in gold coins, and 59% in ‘Deposits’ abroad. (59% of 103.7 tonnes is 61.2 tonnes)
Note the gold deposits with Bank of Nova Scotia and Fortis Bank Bruxelles in 2005 and additionally with the same two banks and with Barclays and Morgan Stanley NY in 2004.
Since the percentage breakdownbetween Romania’s bullionbankdeposits (59%), standardbars (27%) and coins (14%) hasn’t varied much since 2005, and was at a similar mix over various years that I checked such as 2011 and 2014, the conclusion is that Romania has had more than 50% of its gold on constant deposit since at least 2004 (i.e. the original allocated gold is long gone).
The 2005 annual report also states that there were 61 tonnes of Romanian gold stored at the Bank of England. Since Romania had just under 105 tonnes of gold in 2005, this 61 tonnes was referring to the gold deposits, which central banks, as illustrated in numerous other examples, continue to count as their gold even though it has been lent to bullion banks.
Romania therefore had or has 61 tonnes of gold stored at the bank of England.
Note also the reference to central vault, which probably refers to a vault in Bucharest.
The Philippines hold 225 tonnes of gold in its official reserves. In November 2000, when the Bangko Sentral ng Pilipinas (BSP) held 225 tonnes of gold, it explained in a press release titled ‘Shipment of Gold Reserves‘ that it ended up storing 95% of its gold at the Bank of England due to the use of location swaps with a counterparty (probably the BIS) that took delivery of BSP gold, and transferred gold to the BSP account at the Bank of England.
Since 2000, the BSP gold reserves have risen, fallen, and risen again and now total 195 tonnes. Assuming the ‘95% of its gold’ storage arrangement is still in place, then the Philippines has 95% of 195 tonnes, or 185 tonnes stored at the Bank of England.
Greece claims to hold 112.6 tonnes of gold. In 2013, the Greek finance ministry on behalf of the Greek central bank stated that half of Greece’s gold reserves were ‘under custody’ of the Bank of Greece, and the other half was ‘under custody’ of the Federal Reserve Bank of New York (FRBNY), the Bank of England and (very vaguely) Switzerland. Who actually controls Greece’s gold reserves at this point in time is anybody’s guess.
Given that the Federal Reserve Bank of New York was listed by the Greek MinFin as a foreign gold storage location ahead of the Bank of England, the assumption here is that of the 50% of Greece’s gold held abroad, the FRBNY holds more of this portion than the Bank of England. And so the assumption is that the Bank of England holds 40% of the foreign half, i.e. 20% of the total of Greece’s gold, with the FRBNY holding 50% of the foreign half. Taking 112 tonnes of gold as Greece’s total gold holding, 40% of this is 22.4 tonnes stored at the Bank of England. (Note, Greek gold reserves keep increasing incrementally each month by small amounts. As I am not sure what these increases relates to, a recent rounded figure of 112 tonnes has been chosen).
The Banca d’Italia holds 2.451.8 tonnes of gold. Although in 2014, the Banca d’Italia released a document in which it confirmed that some of this gold is held at the Bank of England, there is no evidence to suggest that Italy’s gold in London amounts to more than a few tonnes left over from 1960s transactions.
Bank of England gold set-aside ledgers show that in 1969 there were less than 1000 ‘Good Delivery’ gold bars in the Banca d’Italia gold account at the Bank of England, weighing less than 400,000 ozs in total. This is equal to about 12 tonnes. Most of the Italian gold at the Bank of England was flown back to Rome (and Milan) in the 1960s.
Since there is no public documentation that Banca d’Italia has ever engaged in gold lending (as far as I am aware), then there would be no need for Italy to keep a lot of gold at the Bank of England. Nearly all of Italy’s foreign held gold (over 1,200 tonnes) looks to be in New York (assuming it hasn’t been swapped or used as loan collateral). Italy could have engaged in non-public gold transactions from the Bank of England using gold location swaps from the FRBNY, or from Rome, but there is no evidence of this.
So, this model assumes 12 tonnes of Italian gold is stored at the Bank of England.
Brazil hold 67.2 tonnes of gold reserves. In 2012, Banco Central do Brasil told me by email that all of its gold reserves were in the form of ‘fixed term gold deposits at commercial banks only’. Since the gold would be required to be stored at the Bank of England for these gold deposit transactions to take place, Brazil therefore holds 67.2 tonnes of gold at the Bank of England. See email below:
Banco Central del Ecuador conducted a 3 year gold swap with Goldman Sachs in June 2014 where it swapped 466,000 ozs for US dollar cash This swapped amount of gold has been factored into the World Gold Council data for Ecuador, and the Ecuadorian reserves dropped by 14.5 tonnes in Q2 2014. from 23.28 tonnes to 11.78 tonnes. This swapped amount of 14.5 tonnes is most probably stored at the Bank of England, since Goldman Sachs proposed a similar deal with Venezuela in 2014 where the gold was required to be at the Bank of England for the swap to be initiated.
Bolivia Central de Bolivia holds 42.5 tonnes of gold, all of which is permanently on deposit with bullion banks. The Bolivian Central Bank is very transparent in explaining where its gold is ‘invested’. Hence, it has (until recently) even provided in its financial accounts, the names of the bullion banks which happened to hold its ‘gold deposits’ and the amounts held by each bank.
A recent Banco Central de Bolivia report for 2014 is less revealing and only shows the country distribution of the gold deposits, with 39% in the UK and the rest in France. While this probably refers to the headquarters of the actual bullion banks in question, i.e. Natixis is French etc, it could mean the gold is being attributed to the Bank of England and the Banque de France, so, a conservative approach here is to attribute 39% of 42.5 tonnes to the Bank of England, i.e. 16.6 tonnes stored at the Bank of England.
Peru holds 34.7 tonnes of gold in its official reserves.
At the end of December 2013, Banco Central de Reserva del Peru held 552,191 ounces (17 tonnes) of gold coins which were stored in the Bank’s own vault, and 562,651 troy ounces of “good delivery” gold bars (17.5 tonnes) which were stored in banks abroad, of which 249,702 ounces were in custody and 312,949 ounces in the form of short-term interest bearing deposits. See 2013 annual report.
Since the gold bars are all ‘good delivery’ bars (which is not the case at the FRBNY), and since Peru has still recently been engaging in gold lending, then the evidence suggests that 17.5 tonnes of Peru’s gold is stored at the Bank of England.
Latvia hold 6.62 tonnes of gold in its official reserves after joining the Euro on 1 January 2014 and after transferring just over 1 tonne of gold to the European Central Bank (ECB). All of Latvia’s gold is stored at the Bank of England, therefore Latvia stores 6.62 tonnes of gold at the Bank of England.
Before this transfer of gold to the ECB, Latvia had 248,706 ozs of gold, and it transferred 35,322 ozs to ECB, leaving 213,384 ozs.
The ECB holds 504.8 tonnes of gold. This gold was transferred by the Euro members to the ECB at the launch of the Euro by 1 January 1999. All the ECB gold is de-centrally managed, meaning that it stays where it was when transferred and is still locally ‘managed’ by the bank which transferred that gold to the ECB. Some banks may have transferred gold stored at FRBNY in fulfillment of their requirement, some banks may have transferred gold at the BoE, and countries such as France and Italy may have transferred amounts which are still stored at Banque de France and Banca d’Italia etc. Some of the ECB gold, such as the smaller amount transferred by Latvia, is in the Bank of England. Other amounts of the ECB’s gold are most certainly also at the Bank of England in London.
It would be a separate project to track these transfers. The 1 tonne of Latvian gold transferred to the ECB at the start o 2014 was included in the figures here just as a placeholder, so as to acknowledge that ECB gold is at the Bank of England. Given that the Euro is a competing currency to the US Dollar, the ECB may have more gold than not stored in Europe and not at the Federal Reserve Bank of New York, since ECB gold would logically be safer not stored in the main Reserve Bank of a competing currency bloc.
In its 2014 annual report, the Bank of Iceland said that “The Bank resumed lending gold for investment purposes in June 2014“, and “The Bank loaned gold to foreign financial institutions during the year”.
The Bank of Iceland lent 99.7% of its gold during 2014 because this is the percentage of the gold reserves which are not payable on demand, but are payable in less than 3 months. See below screenshot.
For the purposes of this exercise, Iceland stores 2 tonnes of gold at the Bank of England.
Ghana’s central bank, the Bank of Ghana, holds 8.7 tonnes of gold in its official reserves (precisely 280,872.439 ozs). Of this total, 39.3%, or 3.42 tonnes is held at the Bank of England, with 27.5% at the Federal Reserve Bank of New York, and 29.5% with investment bank UBS. See 2014 annual report.
Interestingly, Ghana refers to its gold account at the Bank of England as a ‘gold set aside’ account, which is the correct name for a Bank of England gold custody account of allocated gold. Probably more interestingly is that most central banks do not use this ‘set aside’ term.
A number of central banks refuse to confirm the location of their gold reserves. I will document this in a future posting. Some of the large holders undoubtedly hold quite a lot of gold at the Bank of England, as do a number of smaller holders. Countries that could fit into this category include Spain, France, Colombia, Lithuania, Sri Lanka, Mauritius, Pakistan, Egypt, Slovenia, Macedonia, Malaysia, Thailand and South Africa. In fact any central bank which has engaged in gold lending is a candidate for having some of its gold stored at the Bank of England.
Spanish people take note. Spain refused to say where its 281.6 tonnes of gold is stored, and Banco de España has the dubious record of being Europe’s least transparent bank as regards gold reserves storage locations. Maybe a project for Spanish journalists.
Banque de France keeps 9% of its 2,435 tonnes of gold reserves abroad, and has in the past engaged in gold lending. So this 9%, or 219 tonnes, is probably stored at the Bank of England.
The ECB and BIS no doubt have more gold stored at the Bank of England than the figures currently reflect. This would also increase the ‘known gold’ total. Egypt is another country which has had a gold set aside account at the Bank of England so is in my view an obvious candidate for the list.
Adding to the known total is therefore a work in progress.
It’s now been 6 months since the LBMA Gold Price auction, the much touted replacement to the London Gold Fixings, was launched on an ICE Benchmark Administration (IBA) platform on Friday 20 March 2015.
For anyone not au fait with the gold price auction, the LBMA Gold Price is a twice daily auction that produces the world’s most widely used gold price benchmark, which is then used as a daily pricing source in gold markets and gold products across the globe.
The 6 month anniversary of the LBMA Gold Price’s launch thus provides an opportune time to revisit a few unresolved and little-noticed aspects of this recently launched auction a.k.a. global benchmark.
Manipulative Behaviour and the FCA
From 1 April 2015, the LBMA Gold Price also became a ‘Regulated Benchmark’ of the UK’s Financial Conduct Authority (FCA) along with 6 other systemically important pricing benchmarks, namely, the LBMA Silver Price, ISDAFix, ICE Brent, WM/Reuters fx, Sonia, and Ronia. These 7 benchmarks join the infamously manipulated LIBOR in now being ‘Regulated Benchmarks’.
Manipulating or attempting to manipulate prices in a Regulated Benchmark is now a criminal offence under the Financial Services Act 2012.
The specifics are set out in Chapter 8 of the FCA’s Market Conduct sourcebook (“MAR”), with the details on ‘identifying potentially manipulative behaviour’ covered in MAR 8.3.6 which says that a benchmark administrator must:
“identify breaches of its practice standards and conduct that may involve manipulation, or attempted manipulation, of the specified benchmark it administers and provide to the oversight committee of the specified benchmark timely updates of suspected breaches of practice standards and attempted manipulation“
“notify the FCA and provide all relevant information where it suspects that, in relation to the specified benchmark it administers, there has been:
(a) a material breach of the benchmark administrator’s practice standards
(b) conduct that may involve manipulation or attempted manipulation of the specified benchmark it administers; or
(c) collusion to manipulate or to attempt to manipulate the specified benchmark it administers.”
and furthermore that the arrangements and procedures referred to above:
“should include (but not be limited to):
(1) carrying out statistical analysis of benchmark submissions, using other relevant market data in order to identify irregularities in benchmark submissions; and
(2) an effective whistle-blowing procedure which allows any person on an anonymous basis to alert the benchmark administrator of conduct that may involve manipulation, or attempted manipulation, of the specified benchmark it administers.”
Section 91 of the UK Financial Services Act 2012 deems it a criminal offence to intentionally engage “in any course of conduct which creates a false or misleading impression as to the price or value of any investment” which creates “an impression may affect the setting of a relevant benchmark”.
Recent Manipulation of Auction Starting Price
All of these FCA rules and the criminalisation of price manipulation offences sound very good in principle.
“4. Findings since go-live: IBA shared with the Committee that:
• IBA, and some direct participants, had observed the price of futures spiking during the minutes immediately before the afternoon gold auction starts.
IBA are now de-emphasising use of the futures as a related market to consider when determining the starting price .”
The fact that IBA has deemed it necessary to follow this course of action (i.e. de-emphasise the use of futures as a starting price determinant), and the fact that some entity or entities have been pushing around futures prices as a means of influencing the LBMA Gold Price starting price suggests that nothing has changed in the gold market since the introduction of the new auction, and that the same players who were actively manipulating the gold price back in 2012 are still doing so, despite this becoming a criminal offence under UK law.
4.12. At the start of the 28 June 2012 Gold Fixing at 3:00 p.m., the Chairman proposed an opening price of USD1,562.00. However, the proposed price quickly dropped to USD1,556.00, following a drop in the price of August COMEX Gold Futures (which was caused by significant selling in the August COMEX Gold Futures market, independent of Barclays and Mr Plunkett).
“4.18. …before the price was fixed, there were a number of further changes in the levels of buying and selling in the 28 June 2012 Gold Fixing, which coincided with an increase in the price of August COMEX Gold Futures.
4.19. As a result of these changes, the level of buying at USD1,558.50 exceeded the level of selling (155 buying/45 selling), and the proposed price was likely to move higher. Given that the price of August COMEX Gold Futures was trading around USD1,560.00 at this time, if the Chairman did move the proposed price in the 28 June 2012 Gold Fixing higher, it was likely to be to a similar price level (which was higher than the Barrier).”
You can read the entire FCA account of the saga of the 28 June 2012 afternoon fixing here, and think about the consequences and meaning of the IBA move to de-emphasis futures prices and what it signals.
Publicly Available Procedures – Not!
Which brings us to the procedures for establishing the auction starting price and subsequent prices for each round of the auction. On 28 April 2015, the IBA LBMA Gold Price web page, under ‘Auction Process’, stated that:
“The chairperson sets the starting price and the price for each round based on publicly available procedures.“
I was interested in reading these publicly available procedures, and learning about the price sources and price hierarchies used within the set of price determinants, so on 28 April 2015, I emailed the IBA communication group and asked:
“I have a question on the LBMA Gold Price methodology.
On the IBA LBMA Gold Price web page (https://www.theice.com/iba/lbma-gold-price) under ‘Auction Process’, point 1 states that “The chairperson sets the starting price and the price for each round based on publicly available procedures“.
Can you direct me to where these ‘publicly available procedures’ are view-able?
Incredibly, IBA received my email that day, and then changed point 1 under ‘Auction Process’ by deleting the original reference to ‘publicly available procedures’ and by copying and pasting in the FAQ answer that I had referred to about ‘in line with current conditions and activity in the auction.”
IBA then responded to my email on the same day, 28 April, without answering the question. The IBA response was:
“Please note the updated text: ‘The chairperson sets the starting price and the price for each round in line with current market conditions and the activity in the auction’. Thank you for pointing this out.“
So, not only did IBA avoid explaining the ‘publicly available procedures‘, they also covered it up and had the cheek to thank me for pointing it out to them. You can see for yourself the reactionary and firefighting tactics used by IBA in perpetuating non-transparency.
Furthermore, the fact that the original web page said that the procedures were publicly available and then they pulled it suggests that at least someone with responsibility in IBA, maybe naively, originally had been of the view that the pricing procedures were to be publicly available.
I emailed IBA again and said:
“This FAQ answer (to the question “How are the prices set for each round of the auction?) doesn’t really explain anything at all.
My question though is, apart from this one line FAQ answer, are there no more in depth ‘publicly available procedures’ available that explain how the opening price is set, what the price sources used are, what pricing hierarchy is used to select an opening price etc..?”
I’ve looked on your web site and in the FAQs and can’t find them. The only brief reference to price determination in the FAQs is that the chairperson”sets the price in line with current market conditions and activity in the auction.”
To which IBA replied:
“This information is not available on our website. However, as you seem to have a few questions, would you be interested in me setting up an off the record briefing with IBA in the next few weeks?”
I did not take IBA up on that offer since I do not think that an off the record briefing is appropriate for something that should be in the public domain. It also highlights the extent to which the vast majority of the financial media are happy to use unidentified sources, off the record briefings, and quotes, and willingly act as the mouthpieces for entities that they are too scared of offending lest they will not get ‘access’ to write their next regurgitated press release for, nor get invited to that entity’s Christmas party.
“‘The names of those selected to oversee ICE’s new gold price benchmarking process will not be disclosed, Finbarr Hutcheson, president of ICE Benchmark Administration (IBA), said.
“We are keeping that anonymous – we don’t think that it’s meaningful to the marketplace to know who’s running that auction and, frankly, the more we kind of feed the story, there’s just going to be more speculation around that,” he said at a briefing at its offices here.
“There’s a legitimate desire to know but actually we don’t want this process to focus on any individual or names of people,” he added.
Not “meaningful to the marketplace to know who’s running the auction“? What sort of statement is that in a free market? If there is a legitimate desire to know, as Hutcheson concedes there is, then why hide the identities?
If anyone needs reminding, the predecessor to the LBMA Gold Price auction was a trading process which, on 23 May 2014, the UK Financial Conduct Authority (FCA) saw fit to fine Barclays £26 million “for failings surrounding the London Gold Fixing.” This was also the first and only precious metals trading process in the UK ever to receive a fine from the FCA.
I would suggest that given the history of a ‘proven to have been manipulated daily gold price auction’, whose successor on launch day primarily consisted of the 4 incumbent participants that comprised the previous Gold Fixings auction (including Barclays), then it certainly is meaningful to the marketplace to know who’s running the new auction.
“’We have a panel of chairpeople that we are going to use and we have internal expertise as well on that, but we are not disclosing the names of those chairmen,’ Hutcheson said. “It will rotate through the panel but we have a significant bench of available external expertise with back-up if you like.”
Hutcheson declined to name how many chairpeople are on the panel.
But if the oversight committee were to feel that it was appropriate for the names to be disclosed, this stance may change, he suggested.”
And why would the oversight committee feel it to be appropriate or not to divulge the names of the chairpersons of the most important gold pricing benchmark in the world?
The Changing of the Guard
Its interesting to see how ICE Benchmark Administration’s description of the chairpersons evolved over a short period after the LBMA Gold Price auction was launched on 20 March.
This was the initial version of the ICE IBA web site description of the Chairperson on 20 March (see screenshot 1 below also):
“The chairperson has extensive experience in the gold market, and is appointed by IBA, and therefore independent of the auction process.”
A week later, a revised, more lengthy version of the Chairperson description had appeared on the ICE IBA web site (see screenshot 2 below also):
“The Chair is appointed by IBA and is independent of any firm associated with the auction, including direct participants. The chair is externally sourced, but works with the IBA team to deliver a robust process for determination of the LBMA Gold Price.”
The Chair facilitates the determination of the LBMA Gold Price by providing his extensive market experience to assist in setting the price in each round of an IBA gold auction.”
By July, the second paragraph of the second version above had been changed to read:
“Both the initial and subsequent round prices are selected by the Chair using their extensive market experience and applied based on an agreed pricing framework.”
So, there is a panel of chairpeople, as Hutcheson told Bulliondesk, who are 4 ‘ex-bankers’ according to Reuters, and who have ‘extensive experience in the gold market’ according to the IBA web site. So these people were previously bankers (which means investment bank staff) who gained their experience of the gold market in investment banks, and who have extensive knowledge of how a gold auction works, and since they are working with London-based IBA on a London-based daily auction, the chairpersons are either London-based or live proximate to London. And finally, according to one of the web site versions above, it’s a ‘He’ or set of ‘Hes’ so we know they are male.
And yet these same people are said to be “independent of any firm associated with the auction, including direct participants.”
Given that there are now 11 direct participants in the LBMA Gold Price auction, namely, Barclays, Bank of China, Goldman Sachs, HSBC Bank USA, JPMorgan Chase Bank, Morgan Stanley, Societe Generale, Bank of Nova Scotia – ScotiaMocatta, Toronto-Dominion Bank, Standard Chartered and UBS, how could ex-bankers based in London with extensive experience of the gold market collectively be independent of all of these banks?
And that’s just the direct participants. What about all the firms associated with the auction, for example, indirect participants who route their auction orders via direct participants?
It would be interesting to hear what IBA and the LBMA define as ‘independent’. Is there any precedent on a definition of ‘independent’ for persons connected to a daily gold auction? Luckily, there is.
“appoint up to two independent qualified individuals to serve on the Committee. A person will be considered to be independent for the purposes of these Terms of Reference if he/she is not, and has not been at any time in the preceding year, an employee or consultant of any Member and does not otherwise have a personal interest in the fixing price or the Fixing Process.”
While this document was referring to a committee whose Members were the directors of the banks running the former auction, at least there is some semblance of a definition of the concept of ‘independent‘ when applied to a gold auction.
So using that yardstick, it would be interesting to measure up the ex-banker chairpersons in the current auction as to how long exactly have they and their handler have been ‘ex’ bankers. Less than a calendar year before 20 March 2015 (i.e. 01 January 2014) would not cut it under a “has not been at any time in the preceding year, an employee or consultant of any Member” test.
And it also begs the question, why is the automated algorithm alluded to by ICE not being used in this LBMA Gold Price auction instead of a human chairperson?
Chairperson description 1
Chairperson description 2
Chairperson description version 3
You will notice from the first description screenshot of the chairperson (above) that on 20 March 2015, ICE IBA stated that:
“Feedback from the market is that the price in the first round of the auction, as well as the prices for the following rounds, is of paramount importance.
As a result, BA has appointed a chairperson from Day 1. In due course, IBA will evaluate developing an algorithm in consultation with the market.“
Then notice that in the second version screenshot about the chairperson, there is no mention of any algorithm. It just vanished.
A slightly different version of the algorithm text appeared in the IBA gold price FAQ document published at launch time:
“Why are you using a Chairperson and not an algorithm for day one?
Feedback from the market is that the setting of the initial price of the first round of the auction, as well as prices for the following rounds, are important. As a result, it is appropriate to have a Chairperson on day one. In due course, IBA will consult on automating the auction process using an automated algorithm.”
A point of information at this juncture. When IBA and LBMA refer to ‘the Market’ they are referring exclusively to LBMA members of the wholesale gold market and not to any of the other hundreds of thousands of global gold market participants who rely on the LBMA Gold Price benchmark as a pricing source. In fact it seems that ‘the Market’ means whatever the LBMA Management Committee decide it means.
It is also worth pointing out that many of the LBMA’s claims on consulting ‘the Market’ are just empty rhetoric, and the consultations are purely for window dressing for decisions that they have already decided on, a case in point being the EY bullion market review commissioned by the LBMA earlier this year that was announced on 27 April and wrapped up by June 2015. This is not too dissimilar to the way FIFA operates, as one correspondent pointed out.
In the case of the above ‘feedback from the market’ about wanting a chairperson, this could very well mean the 4 members of London Gold Market Fixing Limited (LGMFL) who all transitioned from the old auction to the new auction as if nothing had changed. It appears that they did not want anything to change. The old London Gold Fixing with 4 members had a chairperson (most recently Simon Weeks from Scotia) who rotated annually through the directors of (LGMFL), i.e. from Barclays, Scotia Mocatta, HSBC and SocGen.
Finbarr Hutcheson had also referred to this price calculation ‘Algorithm’ on 19 March, the day before the LBMA Gold Price launch. To quote Bulliondesk again:
“The panel of the independent chairs will be responsible for overseeing the process although ICE has indicated that it will be looking to make the process electronic in future.“
The LBMA Silver Price Algorithm
The LBMA Silver Price auction has a separate administrator, Thomson Reuters and a separate platform provider, CME Group. Thomson Reuters has this to say about the opening price on page 8 of its LBMA Silver Price methodology guide:
3.7 Starting Price
The auction platform operator (CME Benchmark Europe Ltd) is responsible for operating the LBMA Silver Price auction, including entering the initial auction price.
The initial auction price value is determined by the auction platform operatorby comparing multiple Market Data sources prior to the auction opening to form a consensus price based on the individual sources of Market Data. The auction platform operator enters the initial auction price before the first round of the auction begins….
For intra-auction prices for each round, the methodology guide says that:
3.8 Manual Price Override
In exceptional circumstances, CME Benchmark Europe Ltd can overrule the automated new price of the next auction round in cases when more significant or finer changes are required. When doing so, the auction platform operator will refer to a composition of live Market Data sources while the auction is in progress.
In the LBMA Silver Price methodology, only the first round is manually input. Subsequent rounds are calculated automatically by the ‘platform’. See page 7 of the guide:
“3.4 End of Round Comparison
[bullet point 2] If the difference between the total buy and sell quantity is greater than the tolerance value, the auction platform determines that the auction is not balanced, automatically cancels orders entered in the auction round by all participants, calculates a new price, and starts a new round with the new price.”
So this is different to the LBMA Gold Price where:
“The chairperson sets the starting price and the price for each round in line with current market conditions and the activity in the auction.”
Six months after the fanfare launch on 20 March 2015, unanswered questions remain:
How robust is the LBMA Gold Price auction mechanism, when within 3 months of launch date, IBA have to tinker with the price sources used to determine the starting price, and de-emphasise one price source due to volatile and seemingly delibrately manipulative futures price movements?
Why does the LBMA Gold Price auction needs a human chairperson throughout the auction and the LBMA Silver Price does not?
What happened to the plans for introducing an algorithm into the auction?
Why have ICE gone to great lengths to prevent the public knowing the identities of the chairpersons?
Why did ICE backtrack on a reference to ‘publicly available procedures‘ that would have explained how the starting price and round prices are determined?
What’s going to happen when the initial six months of the chairpersons’ rotating duties run out on Monday 21 September, as Reuters alluded to back in March?
To that list some further questions could be added:
Where are the Chinese banks ICBC and China Construction, Bank which both expressed interest in becoming direct participants in the LBMA Gold Price auction, going to join?
Where are all the gold mining and gold refining entities that have expressed interest in being direct participants going to join, participants that the ICE auction platform can accommodate right now?
When will the LBMA Gold Price auction move to central clearing on an exchange distinct from LMPCL’s monopoly on clearing predominantly unallocated metal?
When will the prohibitive credit lines enforced by the LBMA be removed as as to allow other non-bank participants to directly participate in the auction without maintaining credit arrangements with the incumbent bullion banks?
These are just some of the questions which financial journalists cannot bring themselves to write about when covering this topic.
Below is an interesting 1988 discussion paper on gold, written by Bank of England staff. The original paper can be found in pdf here. The United Kingdom’s gold reserves belong to HM Treasury (HMT) and are held in the Exchange Equalisation Account (EEA), but managed by the Bank of England on behalf of HMT. The discussion paper is presented with no comment. The introductory context for the paper and the tables and appendices referred to in the paper can be found in the original link as part of the supporting documentation. Discussion paper begins below. Hopefully it improves readability since pdfs are quite awkward to read and, in this case, the pdf text was quite squashed.
THE EXCHANGE EQUALISATION ACCOUNT’S HOLDINGS OF GOLD
The EEA’s holdings of gold, including the gold swapped for ECU’s with the EMCF, amounts to 23.8 mn ozs, or some 740 tons. Since the substantial reduction in stocks in the late 1960’s and early 1970’s resulting from the then balance of payments’ crises, there has been no significant change in holdings. At the annual revaluation of the reserves in March 1988, the gold holdings were worth $8.1 bn. However as a share of total reserves they have fallen significantly from 31 per cent in 1980 to 17 per cent of total spot reserves now. Table 1 shows the change.
Table 2 compares UK official gold holdings since 1969 with those of other industrialised countries and, with Chart 1, shows that, Japan apart, UK holdings are low both in absolute terms and as a proportion of total reserves. As with the UK, there has been little movement in the level of gold holdings by most other countries in recent years.
There are three principal reasons why countries wish to retain an element of gold in their reserves:
(a) as the ultimate store of value, the “war chest” argument;
(b) as an alternative investment to holdings of currencies and other reserve assets;
(c) as an insurance against gold resuming a central role in the international monetary system.
(a) The “war chest” argument
For thousands of years gold has been, in times of war and crisis, the ultimate store of value and medium of exchange. Gold is virtually indestructible, anonymous, mobile and almost universally acceptable. In times of crisis and uncertainty the presence of a sizeable gold holding boosts confidence of creditors, not least because gold is the highest quality asset: unlike foreign currencies, it is not a claim on a debtor (bank or government) and therefore does not have the same risk of default in times of crisis. The war chest argument is still a cogent one but it does not, of course, tell you what size of gold holdings are necessary insurance against catastrophe.
(b) Gold as an investment
The return on gold is measured by the increase in its value over time. This can be compared with the return on alternative financial assets over the same period. Table 3 compares the return on our gold holdings to the return we would have used from holding US Treasury Bills. However, even more than with other financial assets, the profitability of holding gold depends critically upon the time period examined: purchases before 1980 would have been a good investment, purchases subsequently a poor investment because the price peaked in 1980, and since, real interest rates have been positive and global inflation receding.
Table 4 gives average annual rates of return over various time periods when in each case the closing date is 1987 when the average gold price was $487. It uses various starting dates. Those to which greatest weight should be given are probably 1973, since that was the beginning of the period when there was a free market for gold (the unofficial price had been free since 1968, but was probably still constrained by the fixed official price until the period of floating exchange rates began); and 1914, since that was the end of the pre-war period when there were no significant restrictions on gold-holding or prices (of course the nominal price of gold was fixed in most currencies, but since under the gold standard the price level was influenced by the supply of and demand for gold, one can argue that the real price of gold was not constrained in the long-term). For any starting date between 1914 and 1973 the return on gold measured to any date after 1973 is shifted upwards by the effect of the freeing of gold from the fixed nominal official price of $20.67 per ounce until 1934 and $35 per ounce until 1971.
There is a case for extending the starting date backwards to 1879 (when the US adopted the gold standard), but this would include the period of the South African gold discoveries, a geological discovery which is unlikely to be duplicated. Moreover, since the return on gold is being compared to the return on US financial instruments, it may be appropriate to omit the period before 1914, when the US, as a developing, capital-importing country had real interest rates generally above those in London, in contradistinction to the period after the first World War.
Of course, it should also be noted that the past is not necessarily a good guide to the future.
In addition to measuring the average return over different period, it is also important to consider the extent to which the return from gold has fluctuated differently from the return on other assets. Since that may reduce the overall risk of our reserves portfolio if its rate of return is not perfectly correlated with that on the other assets in the portfolio. Whether gold serves such a purpose depends both upon:
(i) how great the variability of its return is relative to that on alternative assets;
(ii) how much less than perfectly correlated is its rate of return with that on other assets.
In fact the answer will depend upon the accounting currency that we are interested in. Assets which reduce risk valued in one currency may not do so when valued in a different currency or basket of currencies.
10. Nevertheless, Annex A shows that over the past 20 years either in dollars or SDR terms, gold has had little or no role to play in reducing risk. Its lack of correlation in return with that on other assets has not been sufficient to outweigh its own relatively high variability of return. The conclusion must be, therefore, that the decision to continue holding significant levels of gold in our reserves must be on other grounds.
(c) Gold in the international monetary system
11. After the collapse of the Bretton Woods system and the abolition of the fixed parities between domestic currencies and gold, the international monetary community decided in principle to remove gold from the international monetary system. Until late 1978 central banks understood not to add to their gold stocks, and the IMF and the USA reduced their gold holdings through auction.
12. However, the move to eliminate gold never gathered much momentum, and has now petered out. The European Monetary System (EMS) has given gold something of a new role in international monetary affairs. Gold deposited with the EMCF can be used to obtain currency through the ECU mobilisation process or (were we in the exchange rate mechanism (ERM) through the use of official ECUs themselves). However, EMS membership (or for that matter, ERM participation) does not constitute an overwhelming ground for maintaining present gold holdings, though ECU mobilisation, within limits, makes gold deposited with the EMCF more liquid.
13. [The mechanism for creating ECUs through the EMCF works on both gold and dollar deposits, so that to the extent that gold was replaced by dollars, the EMCF’s capacity to create ECUs would not be affected. Furthermore, the logic of the system is that an identifiable Community asset is created against the deposit of non-Community reserve assets, namely dollars and gold. There is no reason in logic why EMCF members should not deposit 20 per cent of their reserves in non-Community currencies other than dollars as well, eg. Yen and Swiss Francs, if it is desirable to increase the scale of ECU creation or to compensate for reductions in gold deposits.]
14. Looking forward over the longer term, it is possible that if moves towards more managed rates continue, there will be renewed attempts to restore gold to a more formal role as a reserve asset. It was interesting to note the amount of speculation along these lines in the US last Autumn, following Secretary Baker’s proposal for a commodity indicator (including gold), even though Secretary Baker was at pains to make it clear that he was not advocating a commodity standard. However, the probability of this happening is not sufficiently high to justify the UK retaining its gold stocks on these grounds alone.
15. Gold is therefore left with something more than a residual function. Its possession is likely to increase confidence in the holder’s currency, at least in times of difficulty; and although it is illiquid compared to many financial assets, and expensive to store, in most circumstances it can be used as security for borrowing or to obtain currency through swaps.
The outlook for the gold price
16. Demand and supply in gold are always difficult to measure because of the special nature of the metal. Unlike other commodities, demand for gold does not for the most part lead to “consumption” of the metal; while it may be transmuted into another form it remains as a potential overhang on the market as it can in the right circumstances, such as the price surge in 1980, be re-sold. In addition, the amount of gold coming on to the free market from the communist world in particular is not known, although some estimates are available. Making due allowance for all this, Table 5 attempts to show how supplies of metal coming on to the market in 1985 and 1986 (a period which saw relatively little change in the dollar price) have been absorbed.
17. New gold production has been rising rapidly in recent years. Although production in the main supplier, South Africa, which provides about 50 per cent of total output, has been static, elsewhere it has soared. New technology and the profitability of gold-mining (production costs of $250 per fine ounce against a market price of $439 per fine ounce) have meant that enormous resources have been devoted to the search for gold. Output in Australasia, the USA and Canada in particular has grown rapidly and is expected to continue to do so over the next few years. From 1280 tons in 1986, non-communist output is forecast to rise to 1500 tons in 1990.
18. It is difficult to see where the demand for gold on this scale at the present level of price is going to come from. Investment demand does not seem likely to pick up while inflation remains low and global interest rates remain positive. The increased production in 1986 was only absorbed with little impact on the price by the purchase of some $600 million worth of gold (about 60 tons) by Japan for the Hirohito coin. Without similar large-scale purchases from some source in each of the next few years, pressure on the price must be likely. After 1990, however, increases in supply would, on present forecasts, appear to tail off, while industrial demand in the strong-currency countries, especially Japan, is likely to have increased. In summary, while it is difficult because of supply considerations to be optimistic about the course of gold price over the next few years, thereafter supply and demand may come into better balance.
The transparency of operations in gold
19. The level of the major countries’ gold holdings is readily available, being published each month with a short time lag in IFS, and is potentially the focus of considerable media interest. As stated in paragraph 2 and demonstrated in Table 2, the level of G7 countries’ holdings has altered hardly at all in recent years and the sale or purchase of gold on anything other than a small-scale would rapidly become public and would carry considerable implications for the price. Any move to reduce our present holding would therefore require careful handling if market disruption was to be avoided. Given the present relatively high level of currency reserves, it would be difficult to claim that sales were being made from necessity and thus it would be clear that the action represented a judgement on the future level of the gold price. While this might prove correct in the short-term, it might not do so over a longer period with consequent undesirable publicity. The international consequences also might be unsatisfactory as such action by the UK might have the effect of reopening the debate about the future of gold and freeing other G7 countries from the self-imposed restraint they clearly feel in relation to their holdings.
20. Given the present state of demand and supply mentioned above, any move to increase our present holdings might not have too much impact on the market as a whole. It would, however, run the risk of proving to be an unwise investment and give, on the one hand, comfort to South Africa and the Soviet Union but on the other, as a comment upon the dollar, possibly great offence to the US.
Earnings on gold
21. There are ways in which the EEA can realise a small return on gold, without changing significantly the overall level of stocks, which we do already exploit. Despite the launch of the Britannia the Bank continues to see a small but steady demand for the half sovereign and, to a much smaller degree, the sovereign. These coins are sold at a premium (15 per cent in the case of the half sovereign, 2½ per cent for the sovereign) over their gold content giving rise to modest but useful profits to the EEA. Secondly, within agreed limits, the Bank has also increased its gold lending operations leading to earnings of some £½ million from these activities over the last two years. There may be scope for some further modest increase in this business in the future which will be discussed between the Bank and HMT.
21. (a) the UK’s gold holdings are modest both absolutely and proportionately, as compared with other G7 countries and in relation to the recent past;
(b) gold still has a role as a strategic store of value; as an investment its relative performance depends greatly upon the period over which that performance is measured. It has a small role in our portfolio by helping to diversify risk;
(c) gold holdings also form part of our contribution to the EMCF swap;
(d) demand and supply considerations make us bearish about the price prospects in the short-term although not necessarily in the medium term but in practice there are significant international constraints on our scope for manoeuvre.
23. (a) the EEA’s gold holdings should be retained for the time being, at approximately their present level;
(b) consistent with prudential considerations, we should seek to exploit whatever opportunities there are for earning a return on our gold holdings;
(c) these recommendations should be reviewed at official level if the price of gold moves (up or down) by, say, $100 an ounce from its present level and in the light of developments in the operation of the EMS.
The normally low-key Swiss gold refining market has been thrown into the spotlight with the announcement that private company Valcambi, the world’s largest gold refinery, is being acquired by Indian group Rajesh Exports Ltd (REL), the world’s largest gold jewellery manufacturer.
This acquisition is worth analysing for a number of reasons, namely will the Valcambi-Rajesh transaction impact marginal gold supply out of Switzerland and elsewhere, and how will the transaction, if at all, increase the likelihood of other large gold refineries becoming future acquisition targets?
The announcement of the Valcambi acquisition should not come as a surprise because it was telegraphed in early July by the Economic Times of India. In its article, the Economic Times revealed that Rajesh Exports was in discussions to acquire a large stake in a Swiss gold refinery, and although the identity of the acquiree was not confirmed at that time, the Times said that Rajesh had “sounded out Valcambi…on a possible transaction”.
Since both Rajesh and majority Valcambi shareholder Newmont Mining declined to comment at the time (with Rajesh citing stock exchange rules), the Times and its industry sources were left to speculate that two of the other three large Swiss refineries, Argor-Heraeus or Metalor, might instead be targets, as opposed to Valcambi. Notably, the 4th large Swiss gold refinery, PAMP, was not mentioned in the Economic Times report.
The Times report would suggest that Rajesh Exports took the initiative in searching for a leading precious metals refinery to purchase. However, now that the acquisition has been announced, Rajesh Exports states that it was the Valcambi shareholders who initiated the search for a buyer. In its press release Rajesh states that:
“the owners of Valcambi conducted a global search for divesting Valcambi, after an extensive search selected Rajesh Exports to acquire Valcambi.”
That the search was prolonged was confirmed by India’s Business Standard, which also highlighted that Rajesh Exports was simultaneously on the look-out for a suitor:
“Valcambi shareholders were looking for a buyer for quite some time. We (Rajesh) were also looking to deploy our cash at a safe place, which could generate a fair amount of business interest and help us grow. So, both of us came together and the transaction was concluded.”
But the transaction looks predominantly to have been a strategically planned sale of Valcambi by its holding company European Gold Refineries (i.e. its owners Newmont Mining and a private Swiss investor group), with what looks like input and advice from investment bank Credit Suisse.
A Quick Recap on Valcambi
Before discussing the Valcambi acquisition, its important to understand the Valcambi shareholding structure and the various parties involved with the refinery over its 54 year history.
Balerna based Valcambi was originally incorporated in the southern Swiss Canton of Ticino as Valori & Cambi SA on 15 May 1961, and changed name to Valcambi SA on 30 June 1967. The founders of the original Valori & Cambi, like its successor, seem to have wanted to maintain low profiles, because other than the fact that it was founded by ‘5 Swiss businessmen/entrepreneurs from Mendrisio”, there is little in the public record to identify who these 5 individuals were, since the online company register records don’t so back that far.
In 1967, Credit Suisse bought 50% of the Valcambi refinery, followed by the purchase of another 30% stake in 1968. The final 20% shareholding was purchased in 1980, giving Credit Suisse 100% control of Valcambi from 1980 up to December 2003. In that era, it was not unusual for a large Swiss bank to own a gold refinery, and the other 2 large Swiss banks of the day, UBS and SBC, also owned their own gold refineries (UBS owned Argor and SBC owned Metalor).
In December 2003, some of the same founders of Valcambi (from 1961) joined up with Newmont Mining and established a company called European Gold Refineries SA (EGR), which was 50% owned by Newmont and 50% owned by a group of Swiss investors (whose identities are not easily discernible). EGR then simultaneously bought 100% of Valcambi SA from Credit Suisse, and at the same time acquired a 66.65% shareholding in a company called Finorafa SA, which was a large gold distribution and financier business into the Italian jewellery market.
In their 2003 funding of EGR, Newmont and the Swiss private investor group each put up CHF 15 million in equal combinations of equity and debt.
In early July 2007, Mitsubishi International Corporation (MIC) of Japan bought a 6.55% shareholdings in EGR, with an option to buy a further 26.78% stake by 15 August 2007 (i.e. over 33% in total). Mitsubishi failed to take up its option in August 2007 to buy a larger shareholding in ERG, so this left Newmont and the Swiss investor group each with a shareholding of 46.725%, since their 50% stakes were each reduced by half of the Mitsubishi International Corporation of 6.55%, i.e. reduced by 3.275% each.
Newmont then bought another 15,960 shares in EGR from some of the private investors in April 2008, which increased its stake from 46.725 to 56.67%. This left the Swiss investor group and Mitsubishi holding a combined 43.33%. By this time ERG owned 100% of Finorafa SA as well as 100% of Valcambi, but Finorafa SA was by that time inactive.
Then in mid-November 2008, Mitsubishi had a change of mind and sold its 6.55% stake back to Newmont and the Swiss private investor group. These resold shares seem to have been split fairly equally between Newmont and the private investor group, bringing Newmont’s stake up to 60% By 2009, Finorafa, although owned by EGR, was in liquidation.
For the Valcambi transaction, Rajesh Exports has actually bought European Gold Refineries SA (EGR), which has full ownership of Valcambi SA. To purchase EGR, Rajesh established a Swiss company called Global Gold Refineries AG, which happens to be registered in the Canton of Lucerne (See company register here).
In turn, Global Gold Refineries AG is 95% owned by REL Singapore Pte Ltd, and 5% owned by Rajesh Exports Ltd India (and REL Singapore is fully owned by Rajesh Exports India). See here for the corporate structure of Valcambi and the holding companies. According to Rajesh, REL Singapore was set up primarily to execute international acquisitions and to source gold from mines.
Who were the Swiss Investor Group?
Note that since the acquisition of Valcambi by Rajesh Exports, there are now only 2 directors listed under the Valcambi Board of Directors, namely Valcambi CEO Michael Mesaric, who is staying on as CEO, and new chairman Federico Domenghini. Domenghini is also listed as the only director of the holding company Global Gold Refineries (see above). Interestingly, Michael Mesaric worked in senior roles at Credit Suisse between 1990 and 2002 before joining Valcambi, and is the first of our Credit Suisse connections.
The penultimate board of directors of Valcambi before the acquisition consisted of 6 individuals, 5 of who have now left the board. This penultimate list of directors can be seen here.
Although the full details of the Swiss investors behind Valcambi appear to be hard to find, some potentially relevant facts can be gleaned from the commercial register of the Canton of Ticino and also from the most recent pre-acquisition list of Valcambi board of directors. In addition, Rajesh mentioned one of the main private investors in its stock exchange press release (see below).
European Gold Refineries SA (EGR) was incorporated in Ticino in December 2003. Since 2003, the members of the board of EGR have been a selection of Newmont appointee directors, a selection of Mitsubishi appointees (for a short period), and a handful of other appointees. It is this third group of directors which may provide clues as to who the ‘Swiss private investors’ are, or at least who represents them.
Looking at EGR’s extract from the commercial register, in reverse date order, the most recent directors of EGR representing Newmont Mining (up until late July 2015) were Thomas Mahoney (chairman), Andrew Strelein, and David Farley. In addition, Carlo Camponovo, Luciano Martelli, and Michael Mesaric were listed as directors. Given that Mesaric is the CEO, this leaves Carlo Camponovo and Luciano Martelli as potential representatives of the Swiss investors, because logically, the Swiss private investors would need representatives on the board.
Going back further, ex-directors of Valcambi include Frank Hanagarne, Darren Morcombe, and Pierre Lassonde, all of Newmont, and Haydar Odok and Toshiro Sakai of Mitsubishi. After that we are left with 3 other directors, namely, Davide Camponovo, Emilio Camponovo, and Marco Cavuoto.
From the recent Valcambi board of directors profiles, Luciano Martelli works at Aurofin SA, and is also a director of Aurofin SA. Martelli has in the past also worked at Credit Suisse. Aurofin is a precious metals trading and financing company that was established in 1969 by Emilio Camponovo. Emilio Camponovo is still chairman of Aurofin.
Carlo Camponovo’s Valcambi profile states that he also worked at Credit Suisse from 1993 to 1997, and then worked at Finorafa SA, which is the second company that EGR owned from 2003 until it was liquidated in 2009. Marco Cavuoto was also a director of Finorafa until 2008.
The main reason for illustrating the above is to show the connections between Valcambi, Aurofin, Finorafa, and tangentially Credit Suisse, and also the Camponovo connections. Furthermore, it illustrates the low-key approach that Valcambi seems to have had in specifically naming its private shareholders.
The Valcambi web site even states that “In Switzerland and beyond: our firm deliberately keeps a low profile but has over the years become a key player in the precious metals refining industry” and to prove the point, the quotation is attributed to an unnamed ‘board member’!
Ironically, in the acquisition press release, Rajesh Exports dropped the low-key approach and provided some additional information about the Valcambi shareholders when it mentioned “Mr. Emilio Camponovo” as “the founder and current major share holder of Valcambi“. This suggests that the Camponovos were in the driving seat for the Valcambi sale alongside Newmont (and possibly Credit Suisse as navigator).
Since Valcambi SA and European Gold Refineries SA are both private companies, there is little financial information available about either company. This has even stumped some of Newmont’s sell side analysts on Wall Street, who in their coverage of the sale admit that since Valcambi is a private company, they don’t have much visibility into Newmont’s disposal of Valcambi beyond knowing the net proceeds of the deal.
The Economic Times article on 1 July appears to have had very knowledgeable sources in India since it accurately foresaw that the deal was an all-cash deal for $400 million, 70% of which would be financed from Rajesh’s resources, and the other 30% from “overseas borrowings”.
This was highly prescient, since the announced acquisition turned out to be an all cash deal for $400 million, and Rajesh Exports confirmed at its press conference on 27 July that 30% – 35% of the consideration will be financed by long-term debt (provided by Credit Suisse, no less).
The Rajesh Exports press release states that over the last 3 years, Valcambi booked revenues of US$ 38 billion per annum, and earnings before interest, tax, depreciation and amortization (EBITDA) of US$ 33 million. These revenues look astronomical but they represent the annual average precious metals flows through the refinery being booked at market values (i.e. 945 tonnes of gold and 325 tonnes of silver per annum at market values).
Newmont (the 60% shareholder) will receive net proceeds from the sale of US$119 million. That could mean $200 million net proceeds to the entire shareholder base. Although its unclear as to exactly how much (in net proceeds) the private investor group received. Given that Rajesh is paying $400 million for Valcambi, Rajesh is also taking over or paying down some of the debt of EGR or Valcambi, or else Valcmabi has a quantum of cash on its balance sheet, or both.
Now that the deal has been announced, Newmont has pitched the sale of its stake as a disposal of a non-core asset which it claims will help pay down its debt and focus on its core business. So, being the largest shareholder of Valcambi, and actively wanting to dispose of non-core assets, this reinforces the view that Newmont was the primary driver of the entire ‘global search’ for a buyer of Valcambi.
As mentioned above, Credit Suisse has a long history of involvement with the Valcambi refinery, having fully owned Valcambi from 1980 to 2003. Credit Suisse’s involvement in the new deal also points to ongoing or rekindled relationship with the Swiss private shareholders and Newmont, since it sold the refinery to them in late 2003.
Until 2008, Newmont managed the Valcambi asset through its Merchant Banking group. This group, among other things, took care of “merger and acquisition analysis and negotiations”. Although Newmont’s Merchant Banking group was phased out in 2008, skilled corporate finance individuals at Newmont undoubtedly lent a hand to in the Valcambi disposal project.
Theoretically, Rajesh Exports could have just bought Newmont’s stake in Valcambi and become the new majority shareholder alongside the existing private investors. The fact that they didn’t go down this route could either mean that Rajesh wanted full corporate control, or that the investor group wanted to redeem its investment, or both.
Ramifications of the Valcambi Sale
The sale of the Valcambi refinery now raises questions as to whether its customer base and the mix of destinations for its gold exports from Switzerland will change, and what impact, if any, will the acquisition have on the ability of other countries to acquire Valcambi refined gold.
Rajesh Exports was an existing customer of Valcambi before the acquisition, and probably quite a large Valcambi customer.
“Top Suppliers include Australian Gold Refinery, ANZ Bank and Valcambi Refinery who constitute 90% of total supply of Raw gold to REL“
So Valcambi was already an important supplied to Rajesh. Although Rajesh Exports only consumed about 170 tonnes of gold over its financial year 2014-2015, Rajesh Mehta, chairman of the group stated in his press release that:
“The acquisition is also of national importance for India, as India is the largest consumer of gold in the world, it would be a step in the right direction by an Indian company to own a world-class asset like Valcambi. On a theoretical basis Valcambi is capable of supplying the entire gold requirement of India.“
Gross gold imports (excluding smuggling) into India totals about 750-800 tonnes per annum at the moment. In its 2013 Sustainability Report Valcambi states that its refinery has an annual capacity for gold refining of 1600 tonnes, and a total annual ‘precious metals’ refining capacity of 2000 tonnes. This is what Rajesh Mehta is referring to ‘in theory’ above.
Will Valcambi start supplying all of its output to India? Most probably no. Could this mean that Valcambi will start supplying more of its output to India? Probably yes. Even if it does though, Valcambi still has a lot of spare refinery capacity.
Rajesh Exports seems to have done the Valcambi acquisition for multiple reasons and not just to secure a source of refined gold supply. Rajesh claims that it wants to become a fully integrated major global gold player. (See above link to presentation where Rajesh even had a ‘Mission 2016′ plan to be a ‘fully integrated jewellery company’ by 2016).
Rajesh also had spare cash which it needed to invest in what it referred to as a safe place (i.e. “We were looking to deploy our cash in a safe place” – See Business Standard quote above). And Switzerland remains a universally known ‘safe place’ to deploy cash.
Rajesh already owns some gold mines, and a refinery, as well as gold manufacturing plants, wholesalers and a retailer network of jewellery showrooms which it plans to expand. The Valcambi acquisition allows Rajesh to move back along the gold supply chain. It also presumably will lead to cost savings on acquiring refinery output.
One of the less tangible benefits will be increased information flow about the gold market, both to Rajesh and to Valcambi. Another benefit to Rajesh will be refinery knowledge and skills transfer. Although headquartered in Bangalore in the state of Karnataka in the southwest of India, Rajesh Exports currently has a gold refinery in Uttarakhand in the north of India. This refinery has a gold output of 200 tonnes per annum. Rajesh plans to upgrade this refinery and turn it a subsidiary of Valcambi and then apply for LBMA gold and silver accreditation for the refinery.
One of the main reasons why Valcambi (and its competitors PAMP and Argor-Hereaus) set up in southern Switzerland near the Italian border was that Italy used to be the world’s largest jewellery manufacturer, consuming vast amounts of refined gold as is occurring in present day India. So in some ways, the acquisition of Valcambi by Rajesh Exports Ltd, as the world’s largest gold jewellery manufacturer, is just taking the supply chain logic a step further and going back to the traditional source of the Italian jewellery manufacturers (i.e. Ticino).
All of the above suggest that the acquisition will not end up diverting huge volumes of Valcambi output to India to such an extent that it would impact other customers’ reliance on Valcambi.
Additionally, Valcambi’s CEO, Michael Mesaric said of the deal that “the coming together of REL and Valcambi would ensure that Valcambi improves on it’s global share of gold business, by opening up new markets in India, Middle East and China.” Although Valcambi never broke down its gold exports by destination, about 80% of total Swiss gold exports in 2014 already went to Asia, with India, Hong Kong and China being the top 3 destinations. So what Mesaric is referring to appears to be more of the same, albeit even higher reliance on the existing top export markets.
Furthermore, Valcambi shareholders would not have agreed to the sale to Rajesh if it jeopardised its existing global customer base. Newmont has reiterated its support and will continue to use Valcambi “under the new ownership structure” since it has “long-term contracts with Valcambi for refining the gold produced” from a number of it mines.
In its 2013 sustainability report, Valcambi states that its clients are:
“some of the largest mining companies in the world, premium luxury watch manufacturers,the largest international banks, governments, central banks and scrap dealers”
The report also revealed that on a geographic basis, Valcambi’s ‘business turnover’ was 33% in Europe, 36% in Europe (non EU), 15% in North/South America, 9% in Africa, 4% in Asia, and 3% in Oceania.
Given that the gold exports trade statistics out of Switzerland do not align with the regions of this business turnover data, these figures (which would also include mining company and bullion bank business) must represent where Valcambi books its sales to and/or where the actual clients are based, rather than the ultimately destinations of the refined gold and silver output that are exported from Switzerland,. For example, a London-based bullion bank client of Valcambi that wanted gold refined in Balerna and sent to China would probably be accounted for by Valcambi as a European client, and the China destination of the gold would not get captured in the revenue records.
Valcambi’s refining capacity
Even if Rajesh Exports requires a higher share of the Valcambi refinery output, there is still plenty of spare refinery capacity in the Balerna facility.
Valcambi’s 2013 sustainability report also said that the refinery had an actual ‘product throughput’ of ‘3.8 tons bars and coins per day’ of gold and ‘1.8 tons bars and grain per day’ of silver. Assuming a 5 day week (250 day work year), that would be 950 tonnes of gold throughput and 450 tonnes of silver per annum.
Rajesh Exports just revealed in its press release that over the last 3 years, Valcambi has refined an annual average of 945 tonnes of gold and 325 tonnes of silver (2835 tonnes of gold and 975 tonnes of silver over 3 years). Presumably the last 3 years that Rajesh mentions refers to the last 3 calendar years of 2012-2014.
The London Bullion Market Association (LBMA) doesn’t reveal annual production data of its refinery members on an individual level, however, the LBMA recently published high level totals of the refined gold production of its accredited refiners (LBMA Good Delivery List) over the years 2006 to 2013. What was striking about the data was that total refined gold production of its refinery members reached 6,601 tonnes in 2013, which was 42% higher than total refined gold production in 2012, and also more than double global mine production of 3,016 tonnes of gold in 2013. See table below from LBMA publication:
Total annual refined gold and silver production by LBMA refiners 2006-2013 (tonnes)
So with Valcambi being the largest gold refinery in the world, it would be realistic to suggest that its annual average of 945 tonnes of refined gold output over the last 3 years probably hides the higher refined gold production that it too experienced in 2013 versus 2012. Unfortunately, there is no LBMA 2014 data. Doing a quick hypothetical calculation of Valcambi’s annual gold output over 2012-2014 where 2013 production was 42% higher than 2012, and 2012 production equaled 2014 production, then Valcambi would have refined 828 tonnes of gold in both 2012 and 2014, and a massive 1179 tonnes in 2013. This however would still be below the refinery’s gold output capacity of 1400 tonnes per annum.
So, whichever way you look at it, on average, the Valcambi refinery is not yet running at full capacity for gold, it probably hasn’t ever reached full capacity (even in 2013), and it still has plenty of spare capacity. So even if Rajesh Exports ramps up gold flow from Valcambi to India, other export destinations such as China, South East Asia and the Middle East needn’t suffer as long as mining and bullion bank clients of the refinery can provide metal to make use of the reserve refining capacity.
The other Swiss Gold Refineries
Does the sale of Valcambi foreshadow the sale of any of the other large Swiss gold refineries or increase the likelihood of a similar transaction? I’d say no, but to answer these questions, you may find it helpful to look at the shareholder structure of Valcambi’s competitors in Switzerland, and then decide.
Apart from Valcambi, there are 3 other large gold refineries in Switzerland and 2 smaller refineries. Valcambi’s 3 big competitors are PAMP, Metalor and Argor-Heraeus.
The refineries owned by PAMP and Argor-Heraeus are also located in the south of the Canton of Ticino, literally within walking distance from Valcambi, in what’s known as the golden triangle of gold refineries in the southern tip of Switzerland. As mentioned above, these refineries were established in this area in order to be as near as possible to Milan and the Italian gold industry. Looking at the map below you will see the municipalities of Mendrisio (Argor-Heraeus), Balerna (Valcambi), and Castel San Pietro (PAMP). Balerna is only 4kms from Mendrisio, and 2kms from Castel San Pietro. Notice also the Swiss – Italian border at the bottom of the map south of Chiasso.
Along with Metalor, which is in Marin-Epagnier in the Canton of Neuchâtel in north-west Switzerland, these Big 4 refineries refine the bulk of Switzerland’s (and the world’s) gold. Valcambi, PAMP, Argor-Heraeus and Metalor are all Associates of the LBMA, and PAMP, Argor-Heraeus and Metalor are three of the five refiners on the LBMA’s refiner referee list which helps maintain the LBMA’s Good Delivery System for gold and silver.
Two other smaller companies refine gold in Switzerland in addition to the Big 4. These two companies, also in the Canton of Neuchâtel and located quite close to Metalor, are PX Précinox in La Chaux-de-Fonds, and Cendres + Metaux in Biel. Together they arguably form another golden triangle of refineries, close to the Swiss gold watch industry and incidentally close to the headquarters of the Swiss National Bank in Bern (home of the SNB’s gold vaults and where the BIS’s also stores gold).
The good delivery bars of Valcambi, PAMP, Argor-Heraeus, Metalor and PX Précinox are on the LBMA’s current Good Delivery list for gold, while the bars of Cendres + Metaux are on the LBMA’s former Good Delivery list for gold (transferred to the former list in April 2015).
Because PX Précinox and Cendres + Metaux are smaller than the Big 4, the analysis below only focuses on Metalor, PAMP and Argor Hereaus, all three of which are privately held Swiss companies.
Metalor here refers to Metalor Technologies International SA. Currently the Metalor group is majority owned by French private equity company Astorg Partners SA (www.astorg-partners.com) headquartered in Paris. The remainder of the shares are owned by Swiss individuals and by Metalor management.
The Metalor group is not just a refinery group. It has two others divisions, Advanced Coatings (for electronics and jewellery) and Electrotechnics (silver conductivity electrical contacts used in electrical applications). The refinery division has 4 refineries worldwide, in Neuchatel Switzerland, in the US (North Attleboro, which is south of Boston and is the headquarters of the refining division), in Hong Kong, and in Singapore. The 2012 Metalor annual report states that the group’s refining capacity of fine gold was 650 tonnes per annum in the Swiss, US and Hong Kong refineries. The Singapore refinery was opened in 2013, and since this has a refinery capacity of 150 tonnes, that boosts the total refinery capacity to about 800 tonnes per annum now.
Metalor is the oldest of the Swiss gold refineries and was under the ownership of Swiss Bank Corporation (SBC) from 1918 until 1998. In 1998 a group of Swiss private investors comprising Ernst Thomke, Martin Bisang, Rolf Soiron and Giorgio Behr acquired the majority of shares from UBS. UBS still retained a minority shareholding following this transaction. Thomke then became Metalor chairman until April 2004, after which Bisang was appointed chairman.
Metalor then raised additional capital from another group of Swiss private investors who operated through a British Virgin Islands company called ‘Partners Only’. Zurich business magazine Bilanz speculated as to the identities of these ‘Partners Only’ investors in an article published in 2005, and another published in 2009. These articles list a number of well-known Swiss investors connected to Roche.
In September 2009, Metalor announced that in July 2009, a majority of the private investor shareholders had sold their shareholdings to Astorg Partners SA in an equity funded transaction. The press releases stated that two of the largest investors would invest their proceeds back in with the Astorg transaction, and that Metalor’s management including Scott Morrison, the Metalor CEO, would also become long-term shareholders. One of these 2 ‘largest shareholders’ who stayed on was Martin Bisang (see above). (Metalor press release and Astorg Partners Press Release).
Swiss newspaper NZZ (Neue Zürcher Zeitung) confirmed in 2010 that Belgium headquartered private equity company Sofina had co-invested alongside Astorg Partners, and together they had acquired almost 60% of the shares, which left the remainder of the shares owned by Metalor management as well as Martin Bisang and Daniel Schlatter. Both Bisang and Schlatter are connected to Bellevue Group, a boutique bank in Zurich, owning 20% and 5% of Bellevue shares, respectively. Bellevue actually acted as co-lead financial advisor to Metalor in its sale to Astorg which lists the transaction as spanning 2008-2009. Astorg lists its Metalor investment as being part of its Astorg IV fund.
The board of Metalor now includes Joël Lacourte, Managing Partner of Astorg Partners, Sophie Pochard, Jean-Hubert Vial, and Benjamin Dierickx, all of Astorg Partners, Martin Bisang and Daniel Schlatter of Bellevue Holding AG, and Metalor CEO Scott Morrison. See Neuchâtel company register extract and Bloomberg.
Of the 2008-2009 sale, Martin Bisang has said previously that “it was extremely difficult to find a buyer” for Metalor. This in some ways was because the Lehman induced financial crisis of 2008/2009 impacted transactional values at that time. However, Astorg was looking for acquisition targets in Switzerland at that time, which obviously helped the sale.
Metalor CEO in 2009 Philippe Royer, said that Astorg was a “long-term majority shareholder”. While this is true, private equity companies in most cases eventually want to crystalise their investments, and so its hard to put an exact time-frame on a PE company’s definition of ‘long term’. Maybe 10 years+. The same may be true of the remaining private investors including from Bellevue. A hostile acquirer looking to purchase just the Metalor refineries would have to take on board the other divisions and navigate the complexity of the company. In a similar way a friendly acquirer in the jewellery or investment gold sectors might be put off by the industrial divisions of the group.
Verdict: No change at Metalor in the medium-term.
The Argor-Hereaeus group, located a few minutes drive from Valcambi and PAMP in southern Ticino, has an “annual refining capacity of 450 tonnes for both gold and silver” according to a 2013 company report.
As well as refining, the group produces a range of bars and coins and high precision products for the watch and jewellery sectors.
The current shareholding structure of Argor-Heraeus is quite diverse and consists of parties from three contiguous central European countries, namely, German engineering conglomerate Heraeus, German bank Commerzbank, The Austrian Mint, as well as Argor-Heraeus management. The fragmented shareholder base evolved as follows:
The company, as Argor SA, was established in 1951. Swiss bank Union Bank of Switzerland (UBS) acquired an 80% stake in 1960, and full ownership in 1973. In 1986, Heraeus of Germany purchased a 25% stake from UBS and entered a joint venture with UBS. In 1999 UBS departed leaving Heraeus and the company management with 100% of the shares. Then in April 1999, Commerzbank took a 35% stake, which resulted in Heraeus having 35%, Commerzbank having 35% and Argor-Heraeus management having 30%.
In 2002, the Austrian Mint (owned by the Austrian central bank) acquired a 24.3% interest, which left then Heraeus with 26.5%, Commerzbank with 26.5% and management were said to have 22.7%.
According to the 2013 annual report of the Austrian Mint, it now claims to own 28.6% of the shares of Argor-Heraeus, with an equity value of CHF 122.4 million (and a profit share for 2013 of CHF 19.5 million). According to the 2014 Commerzbank annual report, Commerzbank now owns 31.2% of Argor Heraeus shares with an equity value of CHF 152.7 million (and a 2014 profit share of CHF 22.7 million). In its latest annual report, Heraeus does not reveal its holding in Argor-Heraeus, but if the Austrian Mint and Commerzbank won a combined 59.8%, then that leaves 40.2% for Heraeus and Argor-Heraeus management.
On the website, Heraeus is listed at the top of the shareholder list, so this may indicate that Heraeus has the largest shareholding, which would be above 31%. This would leave management with the remainder.
A complex and diverse shareholder base means a diverse board of directors, and from the Argor-Heraeus SA company registry filing, the board of directors includes, as expected, a cross-section of directors from Commerzbank, the Austrian Mint, and Heraeus, including Gerhard Starsich, CEO and board member of the Austrian Mint, Hans-Jürgen Deutsch of Heraeus Precious Metals, and David Burns, head of commodities at Commerzbank.
All three parties often refer to the strategic benefits of being a shareholder in the Argor-Heraeus refinery so, it seems that the existing formula, whatever it is, is working well.
For example, Commerzbank states that it has a “long-standing cooperation with the refinery Argor-Heraeus S.A. allows us to combine well-founded experience in physical metals with strong expertise in structuring“. Likewise, the Austrian Mint refers to using Argor-Heraeus as a source of refined metal supply, presumably on preferred terms. All parties also presumably get access to information flow about the Swiss gold refining industry and gold demand and supply trends in and out of Switzerland, which is helpful.
In its 2013 annual report, the Austrian Mint said that Argor-Heraeus achieved “large increases in sales and profits in comparison to the preceding year”, so the refinery appears to be a good investment for the various parties also.
It therefore doesn’t seem likely that any of the 3 external shareholders would need to, or want to, dispose of their shareholdings. An acquirer would have to navigate negotiations with a central bank (Austria), a large German bullion bank, and a large German conglomerate, in addition to the Argor-Heraeus management.
Verdict: No change in Argor-Heraeus ownership over the foreseeable future
PAMP (Produits Artistiques Métaux Précieux)
PAMP SA of Castel San Pietro in Ticino, a neighbour of Valcambi and Argor-Heraeus, operates two precious metals refineries, one in Ticino and the other as a joint venture with MMTC in Delhi in India. PAMP SA is fully owned by MKS (Switzerland) Finance SA of Geneva.
Together the two refineries have an annual capacity for 550+ tonnes of gold, and 1200+ tonnes of silver. According to its website, “PAMP handles over 400-metric-tonnes of gold per year”, therefore there is still spare capacity.
MKS, a private company founded in 1979, is actually headquartered in the Netherlands, and has 16 offices around the world. MKS could be described as a physical precious metals refining and distribution company, and also a precious metals trading and financing company. The main office is in Geneva. MKS also owns precious metals bar and coin wholesaler Manfra, Tordella & Brooke (MTB) in New York which will be familiar to some readers as an approved Comex depository for gold. MKS Finance SA is also an Associate of the LBMA.
According to its company registry filing in the Canton of Geneva, the board of MKS (Switzerland) SA includes chairman Marwan Shakarchi, vice-chairman Karma Shakarchi-Liess, Venkata Gopalakrishnan, Hans Isler, Jean-Pierre Roth, and Stanley Walter.
The PAMP SA company filing from Ticino can be seen here.
In India, the PAMP refinery, India’s largest gold and silver refinery, is a joint venture established in 2008 with MMTC, and is known as MMTC-PAMP. MMTC is a ‘Government of India Undertaking’ or Central Public Sector Enterprise (CPSE), and is a huge trading company and the biggest precious metals importer in India. A few of MMTC’s directors are Indian Government appointees and the company’s website even uses a government web site domain (http://mmtclimited.gov.in/).
“MMTC is the largest importer of gold and silver in the Indian sub-continent, handling about 174 MT of gold and 1165 MT of silver during 2011-12. MMTC supplies gold on loan and outright basis to the exporter, bullion dealers and jewellery manufacturers on all India basis.”
MMTC also has its own nationwide retail jewellery showroom network. From an Indian prespective, it’s not surprising that Rajesh Exports would have steered clear of looking to acquire PAMP because of PAMP’s existing relationships with MMTC. Recall that PAMP was not mentioned by the sources quoted by the Economic Times of India as a potential Swiss refinery target, while Valcambi, Metalor and Argor-Heraeus were mentioned. MMTC-PAMP, is the only precious metals refiner in India currently on the LBMA’s good delivery list.
An acquisition of PAMP SA of Switzerland would probably have to be a full acquisition of the entire MKS Finance group becasue PAMP and MKS are closely integrated across a lot of their respective functions. Since MKS seems to be thriving independently, its doubtful if they’d be interested in being taken over. Perhaps they’d be more open to collaboration. Negotiating with one owner as opposed to multiple owners in an acquisition scenario would undoubtedly be easier though.
It’s still unclear though as to how the exact shareholdings of MKS and PAMP are structured. MKS states that it’s a family-owned business and that would mean either exclusive or majority ownership by the founding Shakarchi family. It probably has some management ownership also. But being a private company, its hard to determine if MKS has, or does not have, a set of external private investors.
Verdict: PAMP and MKS will probably remain independent but watch this space
(Update 23/03/2015: The www.goldfixing.com website was permanently switched off in the early morning of Monday 23rd March 2015)
The last ever ‘Gold Fixing’ will take place on the afternoon of Thursday 19th March 2015 at 3.00pm.
Following the last fixing, the www.goldfixing.com website of the London Gold Market Fixing Limited will be immediately and permanently taken off-line as of close of business 19th March (i.e. the web server will be made inaccessible to web browsers).
London Gold Market Fixing Limited (LGMFL) recently confirmed to me that:
“The Gold Fixing website will be taken down completely as of the close of business on Thursday 19-3-15 and from 20-3-15 the new LBMA Gold Price will appear on their website www.lbma.org.uk. All the historical gold Fixing price data on the www.goldfixing.com website is already available on the LBMA website.“
This power-down and switching off of the web server follows a similar manoeuvre on the evening of 14th August 2014, when the web site of the London Silver Market Fixing Limited, www.silverfixing.com, was immediately and permanently switched off (without warning), leaving no trace of the live website.
This is very unusual behavior by the administrators of the fixing web sites and the bullion banks that run the Gold Fixing and Silver Fixing Companies to immediately make the web sites inaccessible. It’s as if the two Fixing Companies want to vanish without a trace from the internet.
The Gold Fixing website domain was first registered on 22 Dec 1999 by firstname.lastname@example.org, and is listed with a tech support contact of email@example.com. See Domain lookup. So there is still a direct reference to Barclays in the web site and in the Sapient app, which is interesting given that Barclays was the firm that was fined by the FCA for manipulating the gold fixing in 2012 and whose trader Daniel Plunkett was also fined for the same offence.
Interestingly, the London Silver Market Fixing Limited has not been wound up, and still exists as a company, and its directors, until recently, represented HSBC, Scotia and Deutsche Bank. The only Deutsche director, New York based Eric Parker, resigned from the company last December. The HSBC and Scotia directors are still in situ.
The London Gold Market Fixing Limited also still exists as a company (obviously), and its directors are representatives of HSBC, Scotia, Barclays and SocGen, and all of these directors are still in situ. The two most recent Deutsche directors, Kevin Rodgers and James Vorley, resigned from the company on 14th May 2014, which was the same day that Deutsche Bank dropped out of the daily Gold Fixing process.
Both the Gold and Silver Fixing Companies have a registered address of c/o Hackwood Secretaries Limited, One Silk Street, London EC2Y 8HQ. Hackwood Secretaries Limited is a company belonging to Linklaters law firm. See the Linklaters ownership of Hackwood Secretaries here. Hackwood Secretaries is also the registered address of London Precious Metal Clearing Limited (LPMCL), the precious metals clearing company of Barclays, HSBC, Scotia, UBS, JP Morgan and Deutsche Bank.
Gold Price Data
From Friday 20th March 2015, the LBMA Gold Price data will appear on the London Bullion Market Association’s (LBMA) web site at http://www.lbma.org.uk/pricing-and-statistics. From 20th March, the daily fixings are being administered by ICE Benchmark Administration (IBA). There is some historical gold fixing price data already on the LBMA site, however it is less detailed than the historical gold price data which had appeared on the www.goldfixing.com web site.
On the Gold Fixing web site, the daily and historical data both included the net volume of gold bars bought and sold at the fixing price, as well as the number of participants who had non-zero interest at the fixing price. This extra detail was added to the gold fixing website in the second half of last year. See screen shots below, both daily and historical. Historical data could also be toggled between dollars, euro and pounds.
In contrast, the gold price data currently displayed on the LBMA web site does not include net volume of bars bought and sold at the fixing price, nor the number of participants declaring a buy or sell interest at the fixing price. See screenshot below.
London Gold Market Fixing Limited’s representative confirmed to me recently that all the historical Gold Fixing price data from www.goldfixing.com is already on the LBMA site. But without the volume and participant data, this claim is not entirely accurate. Adding in the volume (in bars) and the participant totals would make the data more complete.
ICE Benchmark Administration do mention a transparency report which will be published after each auction. This report will contain volume and participant numbers. It remains to be seen if this report will be published on the LBMA website. From the LBMA FAQ:
“At the end of the auction process, IBA will publish the benchmark price. IBA will also publish a Transparency Report showing for each round: the price in USD; the aggregated bid and offer volume; the number of participants; and the timings for each round.”
The current gold price data disclaimer on the LBMA web site (for data up to 19th March) states that “Fixing data reproduced by kind permission of the London Gold Market Fixing Ltd” . Please refer to its website to see licensing requirements for the commercial use of the data as well as the time stamps.”
Incidentally, this existing LBMA disclaimer continues with some very out of date text referring to BBA LIBOR. Now there’s a blast from the past…. “Neither the BBA LIBOR Limited, nor the BBA LIBOR Contributor Banks, nor Reuters, nor the LBMA can be held responsible for any irregularity or inaccuracy of BBA LIBOR (for more details, see “Prices Explained”).”
The new LBMA Gold Price has its own disclaimer which is to be found on the new LBMA Gold Price page, and it reads “LBMA Gold Price (“Benchmark”) is owned by The London Bullion Market Association (“LBMA”), calculated and administered by IBA.”
No Way Back
The reason for highlighting that the Gold Fixing web site is being taken offline is that now there will be no way to access it. This is so because it has not been archived in any meaningful way. This is so because the Gold Fixing website www.goldfixing.com has a ‘Terms’ page that appeared at the entry page which prevented the Internet Archive / Wayback Machine (www.archive.org) from archiving the site’s pages. This means that all of the non-price data on the site disappears when the web site is switched off.
On the Gold Fixing web site, under a section called ‘Policy Documentation’ there were seven documents, including the Code of Conduct of the Gold Fixing company, the Terms of Reference of the gold fixing Supervisory Committee, the names of the people on the Supervisory Committee, and a document retention policy. For those interested, these documents can be found here:
The website also contained an essay about the history of the gold fixing, which I noticed also is on another website here. On the Gold Fixing web site, there had been one small recent addition to this essay to reflect the fact that Deutsche Bank had resigned from the auction. This small addition stated “In May, 2014 Deutsche Bank resigned as a member and the line-up became: Barclays, HSBC, Bank of Nova Scotia Mocatta and Societe Generale.”
The Gold Fixing web site had also prominently displayed the four Members and the most recent chairman, and the logos of the Member banks:
Finally, there was an interesting section explaining “How is the Price Fixed?” which was not really explained very well elsewhere. This will be of relevance to anyone who wants to compare the old price determination process to the new one, and to gauge how much, or how little, the process has changed.
This is especially relevant because ICE Benchmark Administration will be employing a human chairperson in the new LBMA Gold Price auction to determine the opening price and the starting price of each round, as opposed to an algorithm. There will be a panel of chairpersons operating on a rotational basis.
Note that ICE and the LBMA are refusing to divulge the identity of these chairpersons.
ICE claim that this is to ‘preserve the anonymity of the auction. This is a totally bogus and unacceptable reason because without ICE confirming the identity of the chairperson, their claim that the chairperson is fully independent of the participants in the auction cannot be verified and will cause suspicion.
It is also shocking, especially since the identity of the chairperson in the Gold Fixing auctions has always been known, in every fixing from 1919 all the way through to 19th March 2015 (i.e. see the screen shots above).
The financial media should really be asking ‘Who is this Chairperson?’, and reporting on this issue.
According to Reuters, ICE has now said that there will be four chairpersons rotating over a six month period and that these will be ‘four ex-bankers‘. The identities of these four ex-bankers have not been revealed. If anything, this appears to cement the control of the incumbent bullion banks over the entire London Gold Fixing process. Reuters also says that none of the Chinese banks will be participants in the auctions.
How the Price was Fixed
The “How is the Price Fixed?” process from the Gold Fixing site can be read below:
The fixing process is governed by a set of Rules for the Administration and Conduct of The London Gold Market Fixings (the “Fixing Rules”). The current version of the Fixing Rules, made under Article 15.01 of the London Gold Market Fixing Limited’s Articles of Association, became effective on 14 July 2014.
The Company has a Supervisory Committee which is responsible for the oversight of the fixing process. The Company has various policies and procedures to ensure the integrity and quality of the gold fixing price which are available on the Company’s website.
Pursuant to the Fixing Rules, representatives of the four members of the London Gold Market Fixing Limited (the “Company”) dial-in to a secure conference facility to determine the single trading price for gold at 10:30 am and 3:00 pm London time on each London business day.
The fixing process commences with the chairman of the fixing (the “Chair”) determining and announcing the opening price of gold.
(The opening price): The Chair shall identify the opening price. The opening price should be the prevailing US dollar mid-market price for London gold and is identified by the Chair after appropriate consideration of the prevailing spot price and the prevailing bid/offer price in the gold futures market.
(Declaration of interests): Assuming this price, the fixing members aggregate all orders received from clients (both prior to the fix and those received in real-time during the fix) with their own proprietary trading position. Members then declare whether or not they have a net buying or selling interest or if they have no buying or selling interest at the opening price.
(No buying or selling interest): If there is no buying or selling interest, the Chair will announce the trading price as fixed at the opening price. Similarly, if at any point during the fixing process there is no buying or selling interest at a given price, the Chair will announce the price as fixed.
(Only buying or selling interest): If there is only buying or selling interest at the opening price and those buying or selling interests represent more than two of the fixing members (e.g. there are three sellers and one no interest), the Chair will move the opening price higher or lower.
Alternatively, if those buying or selling interests represent two or fewer of the fixing members (e.g. two buyers and two no interests), the Chair will ask the fixing members to indicate the net quantity of gold that they are willing to buy or sell at that price.
Fixing members must declare their net interest in increments of five gold bars and must not declare any interest of less than five bars. If the total quantity offered or wanted is 50 bars or less, the Chair will declare the price as fixed at the opening price. If the total quantity offered or wanted is more than 50 bars, the Chair will move the opening price higher or lower.
Similarly, if at any point during the fixing process there is only buying or selling interest at a given price, the Chair will act as described in paragraphs 8 to 10 above.
(Two way interest:) If there is two-way interest at a given price, the Chair will ask members to indicate the net quantity of gold that they are willing to buy or sell at that price.
If supply meets demand, or the difference between supply and demand is 50 bars or less, the Chair may declare the trading price as fixed. Otherwise, the Chair will progressively move the price up or down in an attempt to meet supply and demand.
An upwards price adjustment will cause (i) the potential fixing price to exceed some purchase order limits, which will have the effect of reducing demand as those orders drop out of the members’ net buying interests; and (ii) the potential fixing price to exceed some sale order limits, which will have the effect of increasing supply as those orders are included in the members’ net selling interests.
A downwards price adjustment will cause (i) the potential fixing price to fall below some purchase order limits, which will have the effect of increasing demand as those orders are included in the members’ net buying interests; and (ii) the potential fixing price to fall below some sale order limits, which will have the effect of decreasing supply as those orders drop out of the members’ net selling interests.
The Chair will repeat this adjustment procedure until supply and demand meet or the imbalance is 50 bars or less and the Chair is able to declare the price as fixed.
(Price increments:) The fixing price must be moved in increments of at least 5 cents and in multiples of five cents during the fixing process, in all case taking account of prevailing market conditions.
The Chair identifies price increments based on an assessment of the current price of gold in the spot and futures markets and the level of buying and selling interests declared in the fixing process.
(The discretion): Where the Chair has been unable to exactly match supply and demand, the fixing members will pro rata the difference between supply and demand amongst themselves.
For example, if there is more buying than selling interest (with two buyers and two sellers) and the difference is 20 bars, each buyer will reduce their buying interest by five bars and each seller will increase their selling interest by five bars. This pro rata arrangement is purely between the fixing members and only affects the amount of gold traded as between those members; it does not affect underlying customer orders.
Where the Chair is unable, through moving the price in increments of 5 cents, to achieve an imbalance of 50 bars or less and three attempts have been made to fix the price at a particular level, the Chair may ask the other fixing members to accept an imbalance of up to 100 bars. All members must agree to this increase.
(Flags): Throughout the fixing process members communicate with their clients who are able to cancel, increase or decrease their interest depending on price changes and the level of buying and selling interest.
If, at any time, a member or one of its clients choose to increase, decrease or withdraw a previously declared buying or selling order, that member may require a short pause to recalculate its net interest. In these circumstances the member may call a “flag” which brings the fixing process to a temporary halt. The gold price cannot be declared by the Chair during such a pause.
The term flag is a reference to when the fixing members would meet in a single place to determine the gold trading price. When a member required a pause they would raise a small flag. The flag would be lowered again when they were ready to proceed with the fixing process.
(Execution): Following the determination of the fixing price, the members will execute trades for gold amongst themselves at 15 cents above the fixing price and in the amounts offered during the fixing process.
The Chair will specify the trades that should be executed between the members. The Fixing Rules specify that the largest seller’s order is filled first by matching that order with the largest buyer’s order, with members’ orders then being matched in descending order size.
Settlement takes place two London/New York business days after the fix. Execution of trades between the members does not affect any arrangements agreed between the members and their clients.
(Determination of the fixing price in euros and pounds sterling): The trading price is published by the Company in three currencies: pounds sterling, euros and US dollars. The fixing process takes place in US dollars. Once the trading price is fixed, the Chair will provide the equivalent trading prices in pounds sterling and euros. The Chair uses the then prevailing exchange rates published on Bloomberg or Reuters for this purpose.
(Publication of the fixing prices): Immediately following a fixing, the Chair posts the fixing price, the time at which the price was fixed, the final buy/sell volume figures on an anonymised basis and the basis on which the price was fixed if the discretion was increased from 50 bars, on the Company’s website. The Chair also sends an email confirmation of the fixing price and the time that the price was fixed to the other members.
The published price is the price for one troy ounce (just over 30 grams) of gold delivered in London in the form of LBMA Good Delivery Bars (approximately 400 troy ounces each).
45 New Bridge Road Singapore059398Singapore Company Registration No.: 201217896Z
Phone: +65 6284 4653