All the rumours and analyses on gold, oil and yuan that are making rounds now in the blogosphere are based on the Nikkei article. But the Nikkei article itself contains zero official sources. Basically, the whole story has been invented by Damon Evans. So, let’s start addressing the claims made in the Nikkei piece.
The only vague connection I could find is that the INE “will accept foreign exchange as … trading margin”. If this includes gold – which technically is not foreign exchange – we will see. In any case, even if gold will be used as trading margin that doesn’t mean the contract is “backed by gold”.
The Nikkei headline clearly reads “China sees new world order withoil benchmark backed by gold”. In this context, the word “backed” for most readers will refer to a fixed parity. In the past, for example, there was a fixed parity between gold and the US dollar; this meant the dollar was backed by gold through the US Treasury; dollars could be redeemed for gold at a fixed price and vice versa. In case of the Nikkei story it would imply a fixed parity between yuan, or oil (this is not clear), and gold. But how would China back anything with gold? Would China’s central bank (the PBOC) defend a fixed price of gold in yuan? And it would do so through an oil futures contract? Impossible.
Quickly ‘the story’ by Nikkei transformed through the blogosphere where analysts suggested the gold in SGE vaults would back the yuan. The problem with this theory is that gold in SGE vaults, (i) isn’t owned by the Chinese government, and (ii) isn’t allowed to be exported from the Chinese domestic market (not very convenient for foreign oil producers). Then analysts suggested the gold in vaults of the Shanghai International Gold Exchange (SGEI) would do the job. But SGEI gold, (i) isn’t owned by the Chinese government either, and (ii) can only have been sourced in the international gold market, payed for with US dollars. So much for the oil-gold trade circumventing US dollars as presented by Nikkei.
Now, let’s zoom in on the logic behind the phrase “crude oil futures contract priced in yuan and convertible into gold”. Futures contracts are an agreement between two traders about the future price of i.e. a commodity (usually denominated in a currency, in the case of the INE contract yuan). There can be no third asset, commodity or currency involved in a futures contract. It cannot be that upon physical delivery of SC – when oil is exchanged for yuan – one of the two traders will say, “you know what, I don’t want yuan (or oil), I want gold”. And, needless to say, the Chinese government will not mingle in the futures trade. The PBOC will not jump in when a SC short or long demands gold. Again, the new INE oil futures contract denominated in yuan will have nothing to do with gold.
Though, be reminded, currently no oil producer is prohibited from buying gold (or something else for that matter) when paid in US dollars. That’s actually the very function of money. Money is used, since ancient times, for what is called indirect exchange. Stuff is sold for money, and with that money all other stuff can be bought. Gold can be bought with the proceeds from oil sales since … forever. An oil futures contract will not suddenly change all that. In the Nikkei piece one analysts was quoted saying:
It’s a transfer of holding their assets in black liquid to yellow metal. It’s a strategic move swapping oil for gold, rather than for U.S. Treasuries, which can be printed out of thin air.
But oil producers are free to buy gold with their moneys (yuan or dollars) with or without the new futures contract. The INE contract will not remove an obligation i.e. for Kuwait to invest in U.S. Treasuries. So, what will change when this new oil-yuan futures contract is launched?
Also bear in mind that futures are hardly ever physically delivered. Futures are used for hedging and speculation. In general, commodities are physically traded in the spot market. Oil for dollars, chocolate for Swiss francs, Dutch cheese for euros, etcetera. Futures contracts are not necessarily needed to sell oil for yuan. Nikkei wrote:
China’s move will allow exporters such as Russia and Iran to circumvent U.S. sanctions by trading in yuan.
But effectively, Venezuela, Russia and Iran can sell their oil to China in exchange for yuan as of this very moment, before the oil-yuan futures contract is live. They also could have done so three years ago. So, in my very humble opinion the new INE contract will not be the instant game changer everybody is talking about.
Perhaps also noteworthy, one commentator on the Nikkei story wrote:
China just announced that any oil-exporter that accepts yuan for oil can convert the oil to gold on the Shanghai Gold Exchange and hedge the hard currency value of the gold on the Shanghai Futures Exchange.
My comments on this paragraph:
As shown above China hasn’t announced anything but an oil-yuan futures contract. Gold has nothing to do with it.
Yuan can technically be spend on gold at the SGE, but gold in the Chinese domestic market (SGE system) is not allowed to be exported. Gold from the SGEI is allowed to be exported but is bought in the international market via yuan with US dollars.
Foreign enterprises, like oil producers, cannot hedge gold on the Shanghai Futures Exchange. The SHFE is not open for international customers. There’s only a spot deferred product listed on the SGE, which is comparable to a futures contract, through which foreign enterprises can hedge gold in yuan. But why would oil producers buy gold and subsequently hedge the metal in yuan. Their end position would be merely exposure to the price of yuan. Why then, not buy a yuan denominated bond with an interest rate? Or hold gold without the hedge?
Prior to publication of the Nikkei article in question I got an email from Evans. He asked me if “China will tie a gold guarantee to the new oil contract?”. I replied, “No. I would be surprised if they did that”. But my quote wasn’t selected for the final publication. The piece only quoted analysts singing the same song. In my view, that’s not what sound journalism is about. First of all Evans didn’t use any official sources, and second he picked analysts that confirmed his bias.
Aside from all the inaccuracies in the Nikkei article, what stands out for me is that indeed a large number of countries is willing to trade oil in yuan and the new INE futures contract is important for this development as it allows oil producers and users to hedge directly in renminbi. And so the INE contract will support oil for yuan trading. That’s what the article should have focussed on.
Although not much has happened yet*, it’s clear Asia wants to get rid of the petrodollar, and it will be interesting to see how this initiative develops.
*Still the majority of global trade is conducted in US dollars, and most foreign exchange reserves are in dollars too. The share of yuan payments, compared to all other currencies, tracked by payment service provider SWIFT were under 2 % in June, down slightly from two years ago. (I have no data on CIPS payments.)
My best estimate as of June 2017 with respect to total above ground gold reserves within the Chinese domestic market is 20,193 tonnes. The majority of these reserves are held by the citizenry, an estimated 16,193 tonnes; the residual 4,000 tonnes, which is a speculative yet conservative estimate, is held by the Chinese central bank the People’s Bank of China.
To substantiate my estimates on above ground gold reserves in China mainland, we’ll first discuss private gold accumulation in China through the Shanghai Gold Exchange (SGE), after which we’ll address official purchases by the People’s Bank of China (PBOC) and its proxies that operate in the international over-the-counter market.
The amount of SGE withdrawals provides a fairly good proxy for Chinese wholesale gold demand, although not all gold passing through the SGE adds to above ground reserves. In China, most scrap supply and disinvestment flows through the Shanghai bourse as well, next to mine output and imports. Needless to say, recycling gold within China doesn’t change the volume of above ground reserves. So, simply using SGE withdrawals won’t fly for calculating above ground reserves. What we’re interested in are net imports and mine production in the Chinese domestic gold market.
There is one region that is importing significant amounts of gold from China, which is Hong Kong, though, this likely isn’t exported from the SFTZ but from the Shenzhen Free Trade Zone. The vast majority of China’s jewellery manufacturers are in Shenzhen, and for quite some years gold jewellery, ornaments, industrial and semi-manufactured parts are being exported from this Chinese fabrication base to Hong Kong. These events haven’t got anything to do with the SGEI in my opinion. Thereby, Hong Kong exports far more gold to China than vice versa.
For computing net gold export from Hong Kong to China we’ll subtract “imports into Hong Kong from China” from “exports and re-exports from Hong Kong to China” (as you know China doesn’t disclose gold trade statistics itself). Imports into Hong Kong accounted for 23 tonnes, while exports and re-exports to China accounted for 333 tonnes. Accordingly, China net imported 311 tonnes from Hong Kong in the first five months of 2017.
If we apply the same math to Switzerland’s customs data, it shows China net imported 172 tonnes from the Swiss in the first six months of this year.
Most definitely Australia has exported gold bullion directly to China in 2017 as well, but the Australian Bureau of Statistics (ABS) has changed its methodology regarding this data somewhere in 2016 and is reluctant to share the details with me. Using my old way to compute Australia’s export directly to China results in 23 tonnes (this number is provisional and will be amended).
The UK, a large gold exporter directly to China in 2014 and 2015, hasn’t shipped any gold directly to China year to date, according to Eurostat.
What’s remarkable is that Chinese true gold demand is far greater than what the World Gold Council (WGC) and GFMS are reporting as “Chinese consumer gold demand”. This is due to incomplete metrics applied by the WGC and GFMS. The immense tonnages imported by China have been waived in previous years, by the aforementioned Western consultancy firms, with dishonest arguments. (If you like to study the details regarding gold demand metrics read this.) In reality, thousands of tonnes are being imported into China and this metal is not coming back in the foreseeable future; causing a bull run on steroids if institutional interest for gold rebounds in the West. Ascending above ground reserves within China imply declining above ground reserves in the rest of the world. And the more scarce the metal in the West, the higher price when demand revives. I’ve described this phenomenon in my previous post How The West Has Been Selling Gold Into A Black Hole. In a forthcoming posts I will add more texture to my analysis.
Domestic mine production in China is not allowed to be exported, effectively all output can be added to above ground reserves. The China Gold Association (CGA) wrote on April 28, 2017, that Chinese domestic mine output in the first quarter accounted for 101 tonnes. Lacking the data for the second quarter, makes me estimate mine production from January until June by doubling 101, which is 202 tonnes. By the way, the CGA added:
Gold is a special product with the dual attribute of general commodity and currency. It is the cornerstone of important global strategic assets and the national financial reserve system. It plays an irreplaceable role in safeguarding national financial stability and economic security.
Based on data publicly available, in the first six months of 2017 China net imported at least 506 tonnes into the domestic market and mined 202 tonnes. An addition of 707 tonnes to Chinese private gold reserves.
Chinese Official Gold Purchases
I can be short on PBOC gold purchases: the Chinese central bank does not buy any gold through the SGE – its increments must be treated in addition to all visible flows – and it buys in secret not to disturb the global market. I’ve shared my analysis regarding the PBOC buying gold through proxies in the international over-the-counter (OTC) market for several years on these pages. Although, my reasoning has been confirmed countless times, it’s worth noting it was affirmed once more not long ago.
Early 2017 world renowned gold analyst Jim Rickards was in a meeting with the three heads of the precious metals trading desks of largest Chinese bullion banks. These gold dealers told Rickards that indeed the PBOC does not buy any gold through the SGE. Rickards stated in the Gold Chronicles podcast published January 17, 2017 (at 25:00) [brackets added by Koos Jansen]:
What I [J. Rickards] don’t know is about the Shanghai Gold Exchange sales, they’re pretty transparent, how much of that is private and how much of that is the government [PBOC]. And I was sort of guessing 50/50, 70/30, whatever. What they told me, and these guys are the dealers [the three heads of the precious metals trading desks], it’s 100 % private. Meaning, the government operates through completely separate channels. The government does not operate through the Shanghai Gold Exchange. … None of what’s going on on the Shanghai Gold Exchange is going to the People’s Bank Of China.
In fact, the PBOC uses Chinese banks as proxies to buy gold in countries like the UK, Switzerland and South-Africa after which the metal is transhipped to Beijing. Note, monetary gold shipments do not show up in customs reports of any country.
I haven’t come across any clues in the past months that have changed my estimate on the PBOC’s true official gold reserves. My best substantiated guess still is 4,000 tonnes (in contrast, the PBOC publicly discloses it holds about 1,840 tonnes). For more information on how and when the PBOC stacked up to 4,000 tonnes, continue reading at the BullionStar Gold University by clicking here.
Estimated Total Gold Reserves China 20,000 Tonnes
Let us put the pieces of the puzzle together. We know the PBOC doesn’t buy gold though the SGE, but prior to 2007 the Chinese gold market wasn’t fully liberalized and back then the PBOC was primary dealer in the domestic market. Any PBOC purchases prior to 2007 could have been from Chinese gold mines. What else do we know? China is said to be a gold importer since the 1990s, suggesting domestically mined gold was not exported after, say, 1994. In the next screen shot from the China Gold Market Report 2010 we can read “China has been a gold importer since the 1990s”.
For the sake of simplicity, we’ll calculate from 1994 onwards. Precious Metals Insights (PMI) has estimated that 2,500 tonnes of gold jewellery were held by the Chinese population in 1994. Furthermore, I have data on Chinese non-monetary gold import starting in 2001 – which started slowly but ramped up in 2010 (exhibit 2).
In 1994 PBOC official reserves accounted for 394 tonnes and Chinese domestic mine output accounted for 90 tonnes. So, our starting point in 1994 is:
From here, we can aggregate domestic mine output and net imports for every succeeding year. As stated above, my assumption is that the PBOC sourced its official gold from domestic mines prior to 2007, but shifted these acquisitions to the international market after 2007. The official gold increments in 2001 (105 tonnes) and 2003 (100 tonnes) I’ve subtracted from “aggregate domestic mine output”, the increments in 2009 (454 tonnes) and onwards I did not subtract from “aggregate domestic mine output”.
The previous calculation has resulted in the following chart:
In the chart the green, blue and grey bars represent private gold reserves, and summed up account for an estimated 16,193 tonnes at the time of writing. The red bars reflect the PBOC’s official gold reserves – I would like to stress this number is speculative – and currently account for 4,000 tonnes. My best estimate as of June 2017 for total above ground gold reserves within the Chinese domestic market is 20,193 tonnes.
Head of the Financial Markets Division of the Dutch central bank, Aerdt Houben, stated in an interview for newspaper Het Financieele Dagblad published in October 2016 that releasing a bar list of the Dutch official gold reserves “would cost hundreds of thousands of euros”. In this post we’ll expose this is virtually impossible – the costs to publish the bar list should be close to zero – and speculate about the far reaching implications of this falsehood.
This story started a couple of years ago. As I am Dutch and concerned not only about my own financial wellbeing but of my country as well, I commenced inquiring my national central bank about the whereabouts and safety of our gold reserves in late 2013. One of my first actions was submitting the local equivalent of a Freedom Of Information Act – in Dutch WOB – to De Nederlandsche Bank (DNB) in order to obtain all written communication of the past decades between DNB and the Federal Reserve Bank Of New York (FRBNY). In 2013 I knew a large share of the Dutch gold was stored at the FRBNY, which I deemed to be an unnecessary risk. In a crisis situation, for example, the US government would be able to confiscate Dutch gold stored on American soil. Unfortunately, DNB responded it’s exempt from certain WOB requests under the banking law from 1998, article 3. (I thought the WOB hit a dead end, though recent developments have changed my mind regarding the legitimacy of the rejection. In a forthcoming post more on my WOB from 2013.)
Subsequently, on 21 November 2014 DNB shocked the financial world by announcing it had covertly repatriated 123 tonnes of gold from the FRBNY vaults. Did DNB question the trustworthiness of the FRBNY like myself? Most likely, as I see few other reasons for repatriating, next to losing trust in the international monetary system itself. The gold wasn’t sold in the Netherlands, as our gold reserves have remained unchanged at 612 tonnes since 2008. Apparently DNB felt safer having less gold stored at the FRBNY. Note, the FRBNY offers institutional clients to store gold free of charge, yet DNB favored to ship it home. From the FRBNY website:
The New York Fed charges account holders a handling fee for gold transactions, including when gold enters or leaves the vault or ownership transfers (moves between compartments), but otherwise does not charge fees for gold storage.
In the press release DNB stated repatriating gold “may have a positive effect on public confidence”. Suggesting the Dutch public – or central bank or government – does not have full faith in the FRBNY as a custodian.
My focus on the Dutch gold, in a way partially mine as our official gold reserves are not owned but merely managed by DNB, was sharpened in 2015. On 26 September of that year I visited the Reinvent Money conference in Rotterdam, the Netherlands. One of the speakers was Jacob De Haan from DNB’s Economics and Research Division. In his presentation, De Haan repeatedly emphasized the importance of transparency in central banking.
Through my WOB experience, however, DNB appeared to be not transparent at all. Thereby, if DNB wants to be transparent and boost public confidence, why doesn’t it publish a gold bar list? The publication of this list would provide one of the most important checks on the existence of the Dutch official gold reserves, as the list can then be cross checked with the inventory lists of gold ETFs and alike, possibly exposing multiple titles of ownership on single gold bars. And this act of transparency could be accomplished within minutes by uploading an excel sheet to the DNB website. When I approached De Haan after the conference and asked why DNB doesn’t put out a gold bar list, he offered me he would look into it. He gave me his email address and we agreed to stay in touch.
Many months pasted, but after countless emails and phone calls DNB finally notified me it would not publish any gold bar list. So much for transparency! The following is what DNB wrote me on 11 August 2016 as the reason not to publish:
…we do not intend to publish a gold bar list. This serves no additional monetary purpose to our aforementioned transparency policy, however it would incur administrative costs.
Administrative costs? There hardly could be administrative costs as this list should be readily available in one or more spreadsheets, I reckoned. When confronting DNB with my logic they replied on 15 August 2016:
DNB has internal gold bar lists, however the conversion of internal lists to documents for publication would create too many administrative burdens.
DNB claims to have “internal lists”, but creating “documents for publication” would create too many administrative burdens. I couldn’t believe it. The only way this excuse would hold was if DNB’s internal lists are non-digital, which then need to be either physically copied or manually inserted in spreadsheet software. However, it’s highly unlikely DNB doesn’t have a digital gold bar list in this day and age. Computers have been widely used since the eighties; that’s more than thirty years ago. One the first applications that computers supported were spreadsheet programs designed for accounting.
Roughly 65 % of the international reserves of the Netherlands are held in gold. Would DNB still keep their precious gold records on pieces of paper?
In my professional opinion the Dutch gold must be meticulously recorded in digital documents and thus publishing a bar list should cost nothing. But showing proof will strengthen my perspective. Up till now this post has been more or less a summary of my previous writings. Down below we’ll zoom in on this material, and reveal why it’s virtually impossible for DNB to gain any administrative burdens for publishing a gold bar list.
Allocated Accounts: These are accounts held by dealers [/custodians] in clients’ names on which are maintained balances of uniquely identifiable bars of metal ‘allocated’ to a specific customer and segregated from other metal held in the vault. The client has full title to this metal with the dealer holding it on the client’s behalf as custodian.
Clients’ holdings will be identified in a weight list of bars showing the unique bar number, gross weight, the assay or fineness of each bar and its fine weight.
Clearly, allocated accounts contain uniquely identifiable gold bars owned by one specific client.
DNB discloses the Dutch official gold reserves position according to the International Monetary Fund’s Balance of Payments and International Investment Position Manual version 6 (BPM6). From DNB [brackets added by Koos Jansen]:
De Nederlandsche Bank [DNB] publishes the balance of payments statistics according to the sixth edition of the Balance of Payments and International Investment Position Manual (BPM6) since October 2014.
The figures for the Netherlands have been adjusted for the period since 2008.
BPM6 forces national authorities to distinguish between gold bullion and unallocated accounts, of which gold bullion can be held in allocated accounts. The German central bank wrote in June 2014 on adopting BPM6 [brackets added by Koos Jansen]:
The new rules are binding for the EU member states [which includes the Netherlands] by virtue of a Council regulation amended by the European Commission.
With regard to reserve assets, gold transactions and positions will in future be subdivided into  gold bullion, which includes gold bars and allocated gold accounts, and  gold receivables, to which no specific gold holdings are assigned [unallocated accounts].
In the next chart we can see the ratio between gold bullion and unallocated accounts of all the Eurosystem’s national central banks. The data has been sourced from the German central bank, as the BundesBank’s website has the most user friendly interface. The Netherlands is said to hold 100 % in gold bullion.
When asked directly, DNB replied all the Dutch official gold is indeed fully allocated. Accordingly, there should be lists from all custodians that show the uniquely identifiable gold bars owned by the Dutch state, as stipulated by LBMA guidelines.
Displayed above in exhibit 1, the Dutch gold is mainly stored abroad. Since November 2014 the breakdown by location is as follows: 31 % in Amsterdam at DNB headquarters, 31 % in New York at the FRBNY, 20 % in Ottawa at the Bank Of Canada (BOC) and 18 % in London at the Bank Of England (BOE).
The BOE And FRBNY Provide Clients A Gold Bar List In Digital Format
We can read the BOE claims to provide clients a digital gold bar list that complies with Annex H of the LBMA’s Specifications for Good Delivery Bars and Application Procedures for Listing, and clients are permitted to inspect their gold at the BOE.
When approached with the same questions, the custodian bank in New York replied it couldn’t comment on this subject. However, there is a bar list of gold stored at the FRBNY in the public domain. For the Gold Reserve Transparency Act (2011, not enacted) the US Treasury published two gold bar lists. The first list in excel sheet format covers the US official gold stored at Fort Knox, Denver and West-Point, which aggregates to 7,715 tonnes (click to download the list). The second list in PDF format covers the US gold stored at the FRBNY, which accounts for 418 tonnes (click to download the list starting on page 128). Below is a screenshot of the FRBNY list:
As shown the FRBNY list fully complies with LBMA standards: included is refinery brand, unique serial/melt number, gross weight, fineness, fine weight and year of manufacturing.
At the bottom of exhibit 8 we read the original document name is “FRBNY Schedule of Inventory of Gold Held.xlsx“. The extension of the document name “.xlsx” means the file was created by Microsoft Excel software, which is the most commonly used spreadsheet application. So, either, the FRBNY keeps its bar lists in excel sheets, or is capable of converting their data to excel format.
Kindly remember the US official gold reserves are owned by the US Treasury, not by the FRBNY. We may conclude the FRBNY is able to provides its clients, such as the US Treasury, gold bar lists in electronic format. There should be no problem whatsoever if DNB would ask the FRBNY for the Dutch gold bar list in excel format.
The Bank of Canada didn’t reply to my inquiries, but it doesn’t matter at this point. It should be clear gold custodians keep their books electronically and fully comply with LBMA standards.
I did find a hint of how the BOC operates. In 1997 Professor Duncan McDowall and his team investigated all gold dealings by the BOC from 1935 until 1956 to evaluate if some of the gold stored in Ottawa had ever been intertwined with Nazi gold. McDowall’s investigation is titled “Due Diligence: A report on the Bank of Canada’s handling of foreign gold during World War II“. One of the professor’s observations with respect to the BOC’s historical documents reads [brackets added by Koos Jansen]:
Fiduciary obligation is similarly represented in the Bank’s [BOC] written dealings with its clients: the entitlement of any client to have a written confirmation of the disposition of the assets they have placed in the care of a bank. A good example of such an obligation in the context of this report would be the regular production of account statements that provided foreign central banks [i.e. DNB] with precise month-end and year-end reckonings of their earmarked gold holdings [allocated accounts] in Ottawa. … Currency Division’s reports on the arrival and departure of gold to and from these accounts therefore provided a meticulous record of foreign clients’ dealings with the Bank.
Even the BOC’s gold books from before the war appeared to be impeccable. I assume the BOC’s current custodial gold bookkeeping is as precise and meticulous now as it was then
DNB Is Likely To Maintain A Gold Bar List in Digital Format
Which leaves us to speculate if DNB itself, as the fourth custodian, holds a digital bar list of the 190 tonnes stored in Amsterdam. Allow me to share why I think they do.
The fact DNB repatriated 123 tonnes in November 2014 from New York, shows they’ve revived their affinity with gold. Few central banks have brought their gold home in recent years, which clearly makes DNB a physical gold advocate. No matter how you look at it, this can’t be denied.
While repatriating DNB took the opportunity to upgrade its vault room at the Frederiksplein in Amsterdam, the Netherlands. Have a look at the DNB gold vault shelving system prior to November 2014 in the picture below:
Now have a look at the new shelving system at the Frederiksplein. This next picture was taken after November 2014:
Obviously, DNB made the structures more robust by switching from wooden shelves to what looks to be iron. DNB consulted the BOE for a new shelving system as the BOE has an identical system since many years prior to 2014. Have a look at a photo from the BOE’s gold vault below:
DNB repatriated 123 tonnes, worth roughly 22 billion euros, from the FRBNY somewhere in the months prior to November 2014, exposing a deep and renewed affinity with gold.
DNB must have received a digital list from New York with the bars transported, as we know the FRBNY keeps its records in an electronic configuration.
While repatriating DNB consulted with the BOE for a robust shelving system in order to upgrade the vault room in Amsterdam, which reaffirms DNB’s careful attention for the gold they store.
Judging from the actions above I dare to say DNB had meticulously, and thus electronically, inventoried the 67 tonnes already stored in Amsterdam before November 2014, or registered this metal when the batch from New York arrived. So very likely all gold stored in Amsterdam is properly recorded in digital format.
A summary of the previous three chapters before we continue:
All the Dutch official gold reserves are held in allocated accounts and thus there are bar lists available, which comply with LBMA standards, from all custodians.
We may conclude all custodians save and distribute their bar lists electronically.
Het Financieele Dagblad
Meanwhile, I was interviewed by Het Fiancieele Dagblad, the Dutch version of the Financial Times, on 27 September 2016 for a weekend special on gold. In the interview I told two FD journalists about my views on gold and my curious encounters with DNB. The next day one of the journalists wrote me he would interview Aerdt Houben, Head of DNB’s Financial Markets Division, for the same gold special and invited me to share what I would ask Houben in his seat. I wrote back I would inquire about the gold bar list and if DNB had ever physically audited all the Dutch gold, among other topics.
FD: Some people are worried the Dutch gold might be gone.
Houben: To a certain degree the people should have trust in us. We are transparent about how much gold we hold and the locations.
FD: Are there any reports and bar lists on this, if so: why aren’t those public?
Houben: The content of the reports is also being checked by our accountants for our annual report. But the gold bar lists that would costs hundreds of thousands of euros. Because many people would have to check the contents and the many updates that are required.
In part Houben said the same as DNB mailed me months before, while specifying the administrative burdens would be several hundreds of thousands of euros. By now we know this is a fallacy.
Regarding the “reports” as mentioned in the FD: according to Houben these “reports” (whatever they are) are checked by DNB’s accountants for the annual report and presumably should proof the existence of the Dutch gold. However, in DNB’s annual report 2016 there is no mentioning of such gold related “reports”, or any gold auditing for that matter. What are these “reports”? And in case these are audit reports, why aren’t those public?
Let’s address the arguments for DNB’s excuse in the FD: “because many people would have to check the contents and the many updates that are required” . This is nonsense. For a proper audit, indeed, the bar lists would have to be checked against the physical inventory at the BOE, FRBNY, BOC and DNB. But, if the Dutch gold is audited by now, what additional checks would have to be done for publishing the bar list? Neither are any “updates” required as everything has been allocated since 2008. All DNB’s justifications have fallen apart.
I asked DNB in November 2016 by email, what exactly are the “reports” mentioned in the FD special, and why can’t DNB publish the gold bar list as provided by the BOE (the one custodian openly stating to provide clients a bar list)? DNB replied [brackets added by Koos Jansen]:
In the red frame it reads:
In response to your messages I can inform you DNB has internal overviews of her gold possessions. These are being checked by external accountants [presumably this means the Dutch gold is audited]. As stated previously, DNB considers publishing a gold bar list to serve no monetary purpose. Thereby, creating a bar list for publication would be costly regarding the different formats delivered by our custodians. This means we will not respect your request for obtaining the gold bar list.
I presume DNB tries to communicate the gold has been audited, but how does one audit gold without a gold bar list that complies with LBMA standards? Only when cross checking bars with an inventory list that discloses all physical characteristics of the bars can audits be performed competently. Bar lists that comply with LBMA standards are indispensable for a physical audit.
Relying on audit documents (“reports“?) drafted by custodians is forgery. A physical audit has to be executed by a third party (not the owner and not the custodian). Common practise in the gold industry is to count 100 % and weigh 2 % of all bars at least once a year for an audit (source Bureau Veritas).
I don’t believe it would take DNB any effort to convert the different list formats by its custodians. It’s all digital and can be converted into one file within seconds. (Though publishing the bar list in different formats is fine too.)
By and by, publishing a gold bar list does serve a monetary purpose as it confirms how much monetary gold as nation truly holds. Without public bar lists countries can more easily create false data.
Sadly, in the email dated 5 January 2017 (exhibit 14) DNB told me it won’t reply to me anymore with respect to their bar list.
Secrets. In the past a central bank was proud of it. Nobody was allowed to know how much gold we had and where it was stored. But the age of central banks cherishing their image of a closed fortress is long gone. Openness is our new policy.
The question is, who’s not telling the truth here? That would be DNB, for sure, and possibly also the BOE and FRBNY.
Just to be clear, the amount of gold leased out by DNB is nil. In 2012 the Dutch Minister Of Finance, De Jager, declared in congress DNB had ceased all gold leasing activities by 2008.
Again, all the Dutch gold is allocated, and yet DNB declared in a newspaper the bar list can’t be published because it would cost “hundreds of thousands of euros“ - this has appeared to be an embarrassing statement andtruly blows DNB’s credibility. If DNB doesn’t wish to disclose its bar list, for whatever reason, it would have done wise not to comment at all on this issue.
But why all the nonsense? Time to speculate. We’ll run through a few scenarios:
Scenario 1) Publishing a bar list might limit DNB’s future flexibility to intervene in financial markets. Currently, DNB hasn’t got any gold leased out. But if the bar list would be published, my central bank would be obstructed in future covert leasing activities.
Suppose, the gold price spikes in five months from now. DNB, or multiple central banks in concert, decide to lease out monetary gold in order to calm the physical market. When the leases would be undone several years later, surely the bars returned will not be the ones lend out. Following this scenario, when a bar list is published now it would be inaccurate in a few years time; showing bars that are long gone, and can show up on private gold ETF inventory lists.
If readers question wether central bankers are capable of ‘not telling the truth’, consider what DNB’s Governor said in an interview early 2012 when asked if he would repatriate any gold from the FRBNY. His answer was firm: “No”. However, shortly after, DNB started to prepare repatriating by reinforcing its headquarters. A new security barrier was constructed around the compound. DNB confirmed to me this was done to prevent any trucks from crashing the building. Likely, the Governor ‘did not tell the truth’ in the interview for strategic reasons.
Scenario 2) It’s possible the BOE claims to provide its clients gold bar lists and auditing rights, but in reality it doesn’t. Meaning, DNB doesn’t have a bar list from the BOE that complies with LBMA standards, which forces them to come up with excuses whenever confronted. This scenario could mean custodial gold at the BOE (and FRBNY) has been embezzled.
In 2016 economist Guillermo Barba pressured the Banco de México to publish a gold bar list of the Mexican gold stored at the BOE. In February 2017 Banco de México delivered Barba a list, but it didn’t satisfy LBMA standards by far. Surely this was done on purpose, because how the list was distributed can never have been how the BOE keeps it. So prior to distribution parts of the list were edited. Barba pressured Banxico once more and received a new list in March 2017 (click here to download the list). But neither did the new list satisfy LBMA standards! The column in the list that reads “serial number“, doesn’t disclose the serial numbers physically inscribed on the bars, which makes them uniquely identifiable, but shows the BOE’s internal numbering. In my opinion Barba was fooled twice by Banxico. Or Banxico was fooled twice by the BOE.
My colleague Ronan Manly tried to obtain a gold bar list from the Irish central bank (CBI); gold stored at the BOE. The CBI’s first response was:
The record concerned does not exist or cannot be found after all reasonable steps to ascertain its whereabouts have been taken, …
Your request was referred to two divisions within the Central Bank of Ireland, … Both divisions have confirmed that they do not hold any such records which fall within the scope of this part of your request. Accordingly, this part of your request is refused.
As far as I know, there has never been a serial number of a gold bar stored at the BOE released in the public domain. It can be the BOE is routinely deceiving its clients by distributing incomplete bar lists.
In the past, the central bank of Austria (OeNB) has failed to audit its gold at the BOE. The Austrian Court of Audit (Der Rechnungshof) wrote in a report in 2015 [brackets added by Koos Jansen]:
… the gold depository contract with the depository in England [BOE] contained deficiencies. With respect to the gold reserves stored abroad, internal auditing measures were lacking.
The OeNB had no appropriate concept to perform audits of its gold reserves. …
Was the OeNB blocked entrance from BOE vaults in 2015?
There is proof FRBNY clients have not been able to audit their gold in New York, at least not in 2007. The German Bundes Rechnungshof released a report in 2012 on the safety of the German gold abroad. Although the report is heavily redacted, on page 10 we read German auditors were not allowed entrance in the FRBNY gold vault to inspect their precious metals, nor were any other clients:
A possibility for the owners to physically record the holdings of their gold is not provided in the terms and conditions. According to the FRBNY, it’s a long-term practice not to allow the owners to inspect their assets in the interest of a safe working and control process. It has confirmed to the Bundesbank that these conditions for gold custody also apply to all other clients that store gold at the FRBNY.
In response to repeated requests from the internal auditors of the Bundesbank, their representatives were given the opportunity to enter the vault system in June 2007 to get an impression of the safety precautions. However, the employees were not given access to the vault compartments, but only to an entrance hall. An examination of gold was therefore not possible.
[Four redacted paragraphs follow]
Clearly the Germans were blocked from auditing their metal, and for decades all FRBNY clients had suffered the same fate.
Not surprisingly, after the developments between the OeNB, BOE, Bundesbank and FRBNY both European central banks decided to repatriate significant shares of their gold stored overseas. And both repatriate over the course of multiple years, which accentuates the friction between the custodians and their clients.
Maybe DNB has experienced the same obstructions in New York as the Germans and hence decided to repatriate.
Scenario 3) DNB just doesn’t feel like publishing a gold bar list.
Who’s to say what the truth is? If readers can think of an additional scenario please comment below.
My final conclusion is that DNB is lying about its gold bar list, which is worrisome as it shouldn’t be necessary, or things behind the scenes are more convoluted and DNB is being lied to by its custodians, which is even more worrisome.
In short, producing a bar list that complies with LBMA standards should be child’s play. And only proper lists can grant us the safety of all the official gold reserves stored at the BOE and FRBNY. As of March 2017 the BOE and FRBNY stored an aggregated 10,821 tonnes of gold, of which the majority is monetary gold.
The Bundesbank, OeNB and DNB all claim their gold is audited by now, but none of them has ever released an audit report. The German central bank wrote me it doesn’t publish its audit reports “since Deutsche Bundesbank and its partners have agreed to maintain confidentiality with regard to the audits”. More secrecy and central bank collusion, no surprises there.
Until central bankers are fully transparent about their gold dealings we can have but mere distrust in them.
This post is part of the Chinese Gold Market essentials series. Click here to go to an overview of all Chinese Gold Market Essentials for a comprehensive understanding of the largest physical gold market globally. This post was updated late 2017.
In this post we will analyze everything there is to learn about PBOC gold purchases. Grasping the exact size of their official gold reserves is unfortunately impossible, assuming they have more than what is publicly disclosed (roughly 1,840 tonnes as of 2017), but there are many clues signalling they’ve covertly bought hundreds if not thousands of tonnes of gold since 2009.
The purpose of this post is to get an overview of all clues and data in order to separate the facts from speculation regarding PBOC purchases. Subsequently, we’ll estimate how much above ground gold is held in China mainland – official (PBOC) and private reserves.
I have been writing for a long time the PBOC does not buy any gold trough the SGE, and therefor PBOC purchases must be seen in addition to the flows of gold going through the famous bourse in Shanghai. Though, it’s necessary to expand on this subject in great detail.
We have a fairly good view on how much gold is going through the SGE and thus how much non-monetary gold is net imported into China from countries like the UK, Switzerland, Hong Kong and Australia (after which it’s not allowed to be exported and thus is accumulated in the mainland). If we add domestic mine supply to imported gold, we can estimate how much gold is held in reserves by the Chinese. But, are any of these visible gold flows bought by the PBOC? Not according to my investigation.
Why The PBOC Does Not Buy Gold Through The SGE
Below are the reasons why I think the PBOC does not buy gold through the SGE.
1) The PBOC prefers to buy gold with US dollars, while all physical gold on the SGE is quoted in yuan.
To get a better grip on this subject it helps if we understand why the PBOC would buy gold in the first place, so let’s sum up all possible incentives. The main objectives for the PBOC to accumulate physical gold are:
Supporting the renminbi for its internationalization (adding trust and credibility to the renminbi).
Owning hard currency (gold) as the cornerstone of capitalism.
Owning reserves (gold) that protect the Chinese economy from external/internal shocks and inflation.
Owning neutral reserves (gold) that are not controlled by a foreign nation (the US).
Diversifying its excessively large US dollar (USD) reserves.
Hedge their USD reserves.
Overthrow the USD hegemony.
After reading this list it should be clear the PBOC rather buys gold with their foreign exchange reserves than with renminbi – China’s FX reserves are worth about $3.0 trillion (2017) and mostly held in USD. The amount of gold currently on the PBOC’s balance sheet (roughly 1,800 tonnes) is disproportionate to the amount of USD held. Hence, the PBOC would prefer to exchange USD for gold. All gold on the SGE is quoted in yuan, meaning the PBOC can’t exchange USD for gold through the SGE. Therefor, the PBOC is more likely to buy gold abroad and these purchases should be added to the visible gold flows we see entering the mainland through the SGE.
2) The PBOC would prefer to buy gold in large 12.5 Kg bars, which are relativly more cheaper. 12.5 Kg bars have almost never traded over the SGE.
It should be said the SGE is a subsidiary of the PBOC. In 2002 China’s central bank erected the SGE to develop the domestic Chinese gold market; for the people to trade gold in yuan. The gold bar sizes available on the SGE are 50 gram, 100 gram, 1 Kg, 3 Kg and 12.5 Kg. Though, the volume of 12.5 Kg contracts (Au99.5 and iAu99.5) ever traded on the SGE is close to nil.
Only the 50g, 100g, 1 Kg and 3 Kg bars are traded, which are consumer sizes. This is a sign the PBOC is not buying gold through the SGE. Gold in large 12.5 Kg bars is relatively cheaper and more attractive for central banks. All central banks, that I know of, hold large bars.
3) The PBOC would prefer to hide its gold purchases.The reason we don’t know how much Chinese official gold reserves are is because this is the best kept secret in China. The PBOC buys gold in utmost secret or it would influence the market and geo-politics. If we think from the PBOC’s point of view, why would they leave a single trace when buying gold? Why would the PBOC buy any gold through the SGE for the world to see? I think they wouldn’t.
The PBOC, having an incentive to exchange its superfluous USD in the international OTC market for gold, is actually obliged to monetize the gold it buys abroad. And when these purchases are transferred to China they will not be disclosed in foreign customs statistics. Subsequently, monetary gold imported into China does not go through the SGE, as only the non-monetary small gold bars go through the SGE.
All visible gold exports to China, traded over the SGE, are not PBOC purchases.
5) The PBOC would prefer to buy gold in an OTC market, not over an exchange like the SGE.The majority of global gold trade is done through the London Bullion Market; the most liquid market there is. This is not a central exchange like the COMEX, but an Over The Counter (OTC) market where buyers and sellers connect (electronically) one on one to trade gold without nosy analysts taking notes. The gold traded can be Loco London – located in London – or elsewhere. The London Bullion Market is ideal for the PBOC, as opposed to the SGE.
6) Another reason for the PBOC to buy abroad would be because it’s cheaper. Gold on the SGE often attracts a significant premium over London spot. Why would the PBOC pay that premium? Especially if it’s buying large quantities.
7) There is anecdotal evidence the PBOC covertly imports gold.Gold industry expert Jim Rickards has written in The Death Of Money (2014):
A senior manager of G4S, one of the world’s leading secure logistics firms, recently revealed to a gold industry executive that he had personally transported gold into China by land through central Asian mountain passes at the head of a column of People’s Liberation Army tanks and armored transport vehicles. This gold was in the form of the 400- ounce “good delivery” bars favored by central banks rather than the smaller one- kilo bars imported through regular channels and favored by retail investors.
This is very interesting. Not only because it demonstrates the PBOC prefers 400 ounce (12.5 Kg) bars over 1 Kg bars, but more so because it confirms the PBOC does not import gold through visible channels. This strengthens my analysis the PBOC does not buy any gold through the SGE. Again, all visible import (in general trade) is required to be sold through the SGE in China.
For information on how monetary gold might be imported into China by the military please read my post China’s Gold Army.
What I [Jim Rickards] don’t know is about the Shanghai Gold Exchange sales, they’re pretty transparent, how much of that is private and how much of that is the government [PBOC]. And I was sort of guessing 50/50, 70/30, whatever. What they told me, and these guys are the dealers [the three heads of the largest bullion banks in China], it’s 100 % private. Meaning, the government operates through completely separate channels. The government does not operate through the Shanghai Gold Exchange. … None of what’s going on on the Shanghai Gold Exchange is going to the People’s Bank Of China.
9) The SGE chairman has stated only consumers buy gold over his exchange. On the LBMA Forum in Singapore on June 25, 2014, a speech was delivered by Xu Luode, then Chairman of the Shanghai Gold Exchange. Below is a snippet from Xu:
Last year, China imported 1,540 tonnes of gold. Such imports, together with the 430 tonnes of gold we produced ourselves, means that we have, in effect, supplied approximately 2,000 tonnes of gold last year.
The 2,000 tonnes of gold were consumed by consumers in China. Of course, we all know that the Chinese ‘dama’ [middle-aged women] accounts for a significant proportion in purchasing gold. So last year, our gold exchange’s inventory reduced by nearly 2,200 tonnes, of which 200 tonnes was recycled gold.
Xu mentions the amount of gold imported into China mainland in 2013 (1,540 tonnes). Would Xu be allowed to break China’s best kept secret on an LBMA forum? Would any of these imports end up at the PBOC? I don’t think so. Moreover, Xu explicitly says all imports and mine output (and scrap supply) has been sold through the SGE system to consumers, not the PBOC.
10) SGE withdrawals are elevated when consumer buying is strong.When examining SGE gold purchases by withdrawals from SGE designated vaults, we can depict a seasonal trend of strong demand around New Year (and in April 2013 and mid 2015). The Chinese people typically buy gold in this period as gifts for each other. Does this trend look like PBOC activity? No.
China’s authorities have a range of options when purchasing gold. They may acquire some of the gold which flows into China; there has been no shortage of that. But there are reasons why they may prefer to buy gold on international markets: gold sold on the SGE is priced in yuan and prospective buyers – for example, the PBoC with large multi-currency reserves – may rather use US dollars than purchasing domestically-priced gold. The international market would have a lot more liquidity too.
12) All reliable sources I have regarding the Chinese gold market tell me the PBOC does not buy gold through the SGE. One of these gentlemen, with ties to bullion banks worldwide, confirmed to me proxies of the PBOC purchase gold directly in the London OTC gold market that is shipped to Beijing with “own airplanes” (possibly by the Chinese gold army). In addition, a Chinese banker told me the PBOC buys gold “in the OTC market”, and, “PBOC proxies can deal directly with refineries in Switzerland and Africa, such as Rand and MKS“.
…none of the 2,200 tonnes of gold was bought by the Chinese central bank. The President said: “The PBOC does not buy gold through the SGE.”
14) The head of a global operations company in security transport leaked in 2013 that 12.5 Kg bars were covertly imported into China for the PBOC. When I interviewed Alex Stanczyk, currently Managing Director of Physical Gold Fund SP, on 9 September 2013 he told me [brackets added by Koos Jansen]:
…We talked to the head of the largest refinery in Switzerland and he told us directly that all that metal that’s coming out of London (904 tons YTD) is being refined into kilo bars and send to China, as well as metal that’s coming in from other areas in the world, that’s all going to China. It’s way more than is being reported or moved through the exchanges. All the kilo bars go to the Chinese people but the PBOC is likely only buying good delivery.
With ‘good delivery’ Alex means the 12.5 Kg large bars that are not being sold through the SGE, but are imported as monetary gold into China without showing up in any country’s customs reports. The quote very clearly indicates that 12.5 Kg bars are imported into China for the PBOC without moving through the SGE.
I shall rest here. The purpose of the listed arguments is to provide you with as much information about the Chinese gold market and PBOC purchases as possible.
In short: according to my analysis the PBOC does not buy gold through the SGE!
How Much Gold Does The PBOC Hold?
What do we know about how much the PBOC has bought? Allow me to present a few clues:
1) From a study by Zhang Bingnan, Vice President of the China Gold Association, we can read the PBOC buys approximately 500 tonnes a year (August 2012):
Forecast the optimal gold reserve capacity in the next 20 years. The conclusion is: 2020, China’s gold optimal reserves should be 5,787 tonnes – 6,750 tonnes. 2030 should be 8,995 tonnes – 10,532 tonnes.
2) Yi Gang, deputy Chinese central bank governor, stated the PBOC is able to buy approximately 500 tonnes a year (March 2013):
We will always keep gold in mind as an option in reserve assets and investments. We are able to import 500-600 tons a year, or more, but we will also take into consideration a stable gold market. If the Chinese government were to buy too much gold, gold prices would surge, a scenario that will hurt Chinese consumers. We can only invest about 1-2 percent of the foreign exchange reserves into gold because the market is too small.
3) Song Xin, President of the China Gold Association, wrote on July 2014 the PBOC should first aim to reach the 4,000 tonnes mark [brackets added by Koos Jansen]:
That is why, in order for gold to fulfill its destined mission, we must raise our [official] gold holdings a great deal, and do so with a solid plan. Step one should take us to the 4,000 tonnes mark, more than Germany and become number two in the world, next, we should increase step by step towards 8,500 tonnes, more than the US.
4) According to Deutsche Bank Markets Research the PBOC buys 500 tonnes a year (November 2014):
In another example, the Chinese government’s open market purchases of roughly 500 tonnes per year have not prevented the gold price from plummeting in recent years.
5) Numerous Chinese analyst suggest the PBOC aims to hold 5,000 tonnes in official gold reserves. Roland Wang, World Gold Council China Managing Director, said (March 26, 2015):
China currently holds about 1.6 percent of its foreign exchange reserves in gold, which is relatively low compared with developed countries and some developing countries, WGC China managing director Roland Wang said.
“The ideal amount should be at least 5 percent of its total forex reserves,” Wang told Reuters in an interview in Hong Kong.
Remarkably, the exact same day Reuters published Wang’s statement Chinese newswire Caixin published a story on gold written by Hedge Fund manager Li Sheng (March 26, 2015):
Gold accounts for only 1.6 percent of China’s forex reserves. This is only a fraction of the figure in the United States and many other developed countries. If China ever increased the level to 5 percent, it would have an enormous impact on global demand for gold.
Li mentions the exact same numbers as Wang from the World Gold Council on the same day: 1.6 % and 5 % of total FX reserves. If China would announce they hold 5 % of total reserves in gold, this would translate into roughly 5,000 tonnes.
If the PBOC would have more than 5,000 tonnes of official gold reserves their ‘gold to GDP ratio‘ would be roughly on par with to the US, Europe and Russia. One of the theories about our current international monetary system – that was detached from gold in 1971 – is that it can shift to a new gold anchored system when all power blocks have equalized the chips (Jim Rickards). In other words, if the US, Europe, Russia and China all have an equal ratio of official gold reserves to their GDP, the international monetary system could make a transition towards gold.
6) Jeremy East, Managing Director Global Head, Metals Trading, Standard Chartered Bank, stated the PBOC is planning to support the renminbi with gold for internationalization (June 25, 2014):
I was at the Shanghai Derivatives Forum at the end of May and one of the speakers was a representative of the [China] Gold Association. He gave us quite an interesting insight into the flavor of what is going on in China from a strategic perspective. Some of the things he talked about included that China planned to change the landscape of world gold markets. He talked about having a strong currency and about having that currency backed by gold, like the US dollar. He also talked about people holding more gold and encouraging more people to hold gold. That is not just individuals, but also the central bank. …
(By the way, China is not planning to “back” their currency with gold in my opinion, they’re more likely to “support” their currency with gold at no fixed parity.)
7) The PBOC could have bought as much as 1,750 tonnes of gold in London in between 2011 and 2015. Although, it’s virtually impossible to track monetary gold flows around the world, as these are exempt from international merchandise trade statistics, the least we can do is try. In September 2015 Ronan Manly and Nick Laird conducted an investigation with respect to how much monetary and non-monetary gold was present in the UK. Luckily, the London Bullion Market Association (LBMA) had published a few estimates in recent years about the total amount of physical gold in London (monetary and non-monetary). In 2011, there were 9,000 tonnes in London. In 2015, there were 6,256 tonnes in London – likely all in 12.5 Kg Good Delivery (GD) bars. These estimates from the LBMA combined with Manly and Laird their investigation have resulted in the next charts (conceived by Nick Laird, Sharelynx):
For a better understanding of physical gold located in London you can read this post by Ronan, this post by Nick or have a look at the next illustration conceived by Jesse from Cafe Americain:
Remarkably, according to estimates by the LBMA the total amount of gold in London decreased by roughly 2,750 tonnes in the period from 2011 until early 2015, while UK’s customs department discloses only 1,000 tonnes were net exported as non-monetary gold during this period. Implying, 1,750 tonnes have been (covertly) exported as monetary gold.
8) Dutch newspaper NRC Handelsblad published an article in 1993 about a 400 tonnes gold sale from the Dutch central bank that was partially bought by the PBOC (you can read a translation of the article here). From NRC:
“With 99 percent certainty we know that the People’s Bank of China has been one of the buyers of the Dutch gold”, said Philip Klapwijk from Goldfields Mining Services, an institute in London affiliated with the South African gold mines that specializes in research into the gold market. Also other London bullion dealers have a strong suspicion that China was involved in the gold sales of DNB. “We have noted that the Chinese central bank has bought gold in recent months”, said John Coley of the London bullion dealer Sharp Pixley and spokesman of the London Bullion Market Association.
I should add, in the nineties the PBOC was the primary (monopoly) dealer in the Chinese domestic gold market and in theory could have sold the gold to Chinese jewelry fabricators.
Facts And Speculation
Let’s chew on some numbers. In the first chart below I’ve plotted a conservative estimate of the total above ground gold reserves in China mainland in June 2017. This conservative estimate is based on the assumption the PBOC owned 1,842 tonnes in June 2015 (this is what the PBOC officially discloses), and the rest of imports and mining were added to private holdings.
What about the amount of private reserves displayed in the chart? Let me explain my calculations from the starting point in 1994. Precious Metals Insights (PMI) has estimated that 2,500 tonnes of gold where held by population in the mainland in 1994, that’s the dark grey jewelry base you can see in the chart. The PBOC official reserves in 1994 accounted for 394 tonnes. In addition, Chinese domestic mining was 90 tonnes in 1994. Below is a chart showing historic Chinese domestic mining output.
China is said to be a gold “importer since the 1990s”, suggesting domestically mined gold was not exported after 1994. In the next screen shot from the China Gold Market Report 2010 we can read “China has been a gold importer since the 1990s”.
Furthermore, China began importing (non-monetary) gold a few years ago, have a look the next chart that shows historic gold trade between Hong Kong and China. Net imports ramp up in 2010. Other countries than Hong Kong, such as Switzerland, also started to visibly export to China after 2010.
So, the starting point in the first chart on Estimated Total Gold Reserves in 1994 in China mainland is computed as:
Subsequently, I’ve added annually domestic mining output – as the Chinese didn’t net export any gold since 1994 – and net imports every succeeding year in the chart. I’ve subtracted all PBOC official reserves gains before 2007 from cumulative domestic mining, the gains after 2007 have I have not subtracted from cumulative domestic mining. Why? Two reasons:
The Chinese domestic gold market (SGE) was fully liberalized in 2007 after which I think the PBOC stopped taking in any gold from domestic mines. (Prior to the liberalisation of the Chinese domestic gold market the PBOC was the primary dealer.)
According to my calculations, from 2003 until 2009 total Chinese gold supply (scrap + mine + import from Hong Kong) wasn’t sufficient to meet total consumer demand and 454 tonnes in PBOC purchases over that period. Although the PBOC claims all purchases before 2009 were done from domestic mines and scrap, I don’t think that’s possible. Hence, I think the PBOC started invisible import somewhere in between 2003 and 2009. And therefor, anything the PBOC added to its official gold reserves after 2007 I did not subtract from cumulative domestic mining.
Lead by the aforementioned calculations, the PBOC had accumulated 1,843 tonnes and the Chinese private sector 16,193 tonnes in June 2017 2016 (in total 18,036 tonnes).This does not capture gold in the black market, that thrived before 2002, neither any assets from wealthy Chinese families. It’s the most conservative estimate I can make using all data I could find.
However, in my opinion the PBOC has bought a lot more gold in recent years. What all the clues mentioned above have in common, is that the PBOC has bought roughly 500 tonnes of gold per year since 2009. Let’s make a new more speculative chart:
The above chart is a copy of the previous conservative estimate, supplemented by 500 tonnes per year since 2009 in PBOC purchases, which I have not subtracted from cumulative domestic mining or cumulative import, as my assumption is this gold has invisibly been imported and not bought through the SGE. (I stopped adding 500 tonnes per year after 2015 because evidence is lacking.)
Speculating: the PBOC has accumulated at least 4,000 tonnes and the Chinese private sector 16,193 tonnes as of late June 2017 (in total 20,193 tonnes).
Above you could read clues from Song Xin (China Gold Association, July, 2014) and Jeremy East (June 25, 2014) about China working on a ‘new monetary system’ that will include gold. Something similar was said by Zhou Ming, General Manager of the Precious Metals Department at ICBC, when Jeremy East asked him at the LBMA forum in Singapore (June, 2014) if the statement “Western gold moves East” was true:
With the status of the US dollar as the international reserve currency is shaky, a new global currency setup is being conceived. Uncertain changes will happen to gold’s traditional dollar-pricing so the US dollar’s influence on gold pricing needs to be re-evaluated.
But, of course, this could also relate to Ming bragging about renminbi internationalisation.
In response to a FOIA request the US Mint has finally released reports drafted from 1993 through 2008 related to the physical audits of the US official gold reserves. However, the documents released are incomplete and reveal the audit procedures have not been executed proficiently. Moreover, because the Mint could not honor its promises in full the costs ($3,144.96 US dollars) of the FOIA request have been refunded.
Thanks to my readers that donated to the crowdfunding campaign I’ve been able to force the US Mint through a Freedom Of Information Act (FOIA) request to hand over documents related to the physical audits of the US official gold reserves stored at the Mint; also referred to as Deep Storage gold. Although the PDF-package digitally sent to me is redacted, incomplete, includes pages copied twice and materials I didn’t ask for, it’s the closest thing that I’ve ever seen to physical audit documentation of gold at Fort Knox and the other Mint depositories drafted in between 1993 and 2008.
What is worrying is that the reports now in my possession reveal the audit procedures have not competently been executed. Combine that with the fact the documents are incomplete and redacted, and the result is suspicion of fraud. In this blog post we’ll have a first critical look at the reports and the problems to be found within.
For starters, allow me to expand on what I think happened at the Mint’s headquarter on the 8th floor at 801 9th Street NW Washington DC, before these documents were sent to me.
It should be clear that the US Treasury (owner of the gold), US Mint (main custodian), Federal Reserve Bank Of New York (second custodian), and the Office Inspector General of the US Treasury (head auditor), are reluctant to disclose information about the audits of the gold at the four largest depositories that store over 8,000 fine metric tonnes. Consider that the most seasoned gold analysts aren’t even aware this gold is audited.
What nobody knows is that according the US government 100 per cent of the Deep Storage gold has been audited in between 1974 and 2008 (page 4). This period can be divided in two chapters: the first runs from 1974 until 1986 when the Committee for Continuing Audit of the U.S. Government-owned Goldverified the majority of the Deep Storage metal. The second chapter covers 1993 until 2008 when the residual was examined under the supervision of the Office Inspector General of the US Treasury. In my previous posts on this subject we focused on the first chapter, what is written below skims the surface of the second. As promised, eventually I will publish a full in-depth analysis of all chapters (there are additional chapters in the fifties, from 1986-1993, in 2009, 2010 and 2011).
Over the years my inquiries at the US government though regular channels have produced little intelligence about the physical audits of the Deep Storage gold. Some departments cooperated at first, but eventually they stopped replying emails or just hang up the phone while I was talking. The second layer of defense was raised when I started submitting FOIAs. Instead of honoring my requests they tried to delay and dodge most appeals. Clearly, the US government prefers not to answer my questions than to flaunt with the audit results.
However, in 2016 I embraced the motivation to push through and find out how many gold bars were counted, weighed and assayed in between 1993 and 2008, when allegedly the last series of physical audits was conducted. Not surprisingly, zero US government departments could provide me the information I was looking for, but through certain FOIAs I obtained leads to submit new FOIAs, and so on 12 Augustus 2016 I demanded, inter alia, the “memoranda submitted by the US Mint Director’s representative regarding audits of the Mint Schedule of Custodial Gold and Silver Reserves to the Chief Financial Officer drafted from 1993 through 2008”. The Mint replied this request would costs me $3,144.96 dollars because it would take 40 hours to search the respective documents, 8 hours for review, and additional costs would be incurred to duplicate 1,200 pages. I thought this was hogwash – 1,200 pages seemed out of proportion for such memoranda, how hard can it be to find a few pages and how did they know it were going to be 1,200 pages if they had to search 40 hours for it – but decided to start a crowdfunding campaign to collect the money.
Within 24 hours the campaign was completed and late August 2016 I sent the Mint a check, in the hopes to receive the documents a.s.a.p.. After the Mint pretended the check was missing for a few weeks, they communicated on 28 September 2016 the funds had arrived and they were working to get the requested documents out to me (exhibit 1).
Months past but nothing happened. I sent several emails and called the Mint three times, but time and time again I was maintained with false excuses. Then, finally, on 23 December 2016 the Mint delivered the documents I paid for. Sort of. Instead of 1,200 pages I received 223 redacted pages that contained 68 pages of reports I didn’t ask for and 21 pages that were copied twice. Effectively, I got 134 pages related to my FOIA request.
When I confronted the Mint I paid $3,144.96 dollars for a meager 134 pages they agreed the costs had been estimated to high and a refund was reasonable. Actually, they told me they never cashed the check. So, quickly I told my bank to cancel the check and ordered my crowdfunding platform to refund all my donors.
As of now all donors to my crowdfunding campaign should have received their money back (if not, please write me an email, see below for my address). From the bottom of my heart I would like to thank everyone for the loan that made this operation possible1!
For me a slight doubt remained if the Mint had tried to fend me off by asking a disproportionate amount of money for a few pages that I assume are alphabetically archived, or that they handled my case in all honesty. A skeptical mind would think the former. To find out I read the internal emails of the Mint employees that handled my FOIA. Those are not directly publicly available, but I was told a trick by more experienced FOIA scholars that reached out to me after I published my previous blog posts on this subject, to ask the Mint for internal emails through, what else, a Freedom Of Information Act request (exhibit 2).
And it worked! On 10 January 2017 I received all (I hope) emails from the Mint I was looking for. Including one wherein Audit Liaison at the United States Mint Tom Noziglia makes an estimate for the costs of my FOIA request of 12 August 2016. Read below (exhibit 3).
At first sight it seems Noziglia and his office stick to prudent protocols. But possibly this email is a veil, meant to deceive me if I would ever read it. Actually, yes, I think it’s a cloak and I’ll share my theory.
We can read from Noziglia, “as Audit Liaison at the US Mint, I [Noziglia] am responsible for the coordination of all external audit initiatives … I have extensive experience in precious metal inventory, … I … coordinate the execution of the annual OIG [Office Inspector General] Joint Seal Inspection of the Custodial Gold at the US Mint”. This page tells us Noziglia is one of the auditors of the US official gold reserves. So, the email above (exhibit 3) was written by the auditor who was involved in the procedures of which I requested the documentation. Noziglia must have known my inquiry could be simply honored by sending just a few pages of documentation, as he was a co-author of the documents in question.
Firstly, with the benefit of hindsight we know Noziglia was lying in his email because by now I have the documents that count only 134 pages, and he was the coordinator of the annual inspections of custodial gold at the Mint. He must have known there were no “1,200 pages in 80 boxes” and so his $2640.00 dollar estimate is a hoax. I think Noziglia wrote the email expecting I would NOT pay the ludicrous amount of dollars, but possibly DID submit a new FOIA to view the Mint’s internal emails. Chances are slim someone could pay $3,144.96 dollars right? But I’m not the first who submits an additional FOIA to obtain internal emails. Hundreds of people went before me, this is a well-known trick for FOIA pundits, and many public servants in the US must be aware of this hazard. Hence I reckon public servants consciously write emails to colleagues, as if these will be publicly released some day. I’ve come to understand submitting and answering FOIAs is nothing but a cat and mouse game.
Second, the Mint never cashed the check. If they really thought they would have to search 40 hours, why not cash the check immediately and get busy? I guess they knew very well there was no searching required.
Third, in case Noziglia had never seen a “memoranda submitted by the US Mint Director’s representative regarding audits of the Mint Schedule of Custodial Gold and Silver Reserves to the Chief Financial Officer”, which is not likely but let’s give him the benefit of the doubt, he could have viewed the most recent version at his office that wasn’t sent to the National Archives (NARA) yet. By doing so he would have learned very effectively these annual memoranda count only a few pages.
Fourth, Noziglia states in his email (exhibit 3) he’s not sure if he will find the documents at all. But this is impossible because he’s a dedicated Mint auditor so he must know what documents the Mint sends to NARA every year. In addition, there was no need for Noziglia to “order off site” boxes, because he simply could have commanded NARA staff to deliver specific documents – this is common practice.
Fifth, in the CC of Noziglia’s email is Kenyatta Fletcher, who is the Chief of the Accounting Division of the Mint. If, which is a big if, Noziglia didn’t know what I was looking for, Fletcher would’ve known these documents wouldn’t count 1,200 pages. But still I was charged a laughable $3,144.96 dollars.
Sixth, Noziglia’s estimate is $2.640.00 dollars, but I have no emails that clarify why $504.96 dollars were added for a total of $3,144.96 dollars I was charged. This indicates, Mint staff communicated in person or through phone calls to finalize my request, and so could have done likewise to handle it in general. Concluding, Noziglia’s email doesn’t paint the full picture of the internel communication.
Seventh, please read what Noziglia’s colleague Grimsby replied to him after 4 minutes.
“Great email”? Why would Grimsby praise Noziglia for his email? If Grimsby would have written,“I agree”, I can understand. But,“great email”? Perhaps Grimsby meant to write, “great calculation that makes no sense, but is likely deceive an ignorant FOIA requester if he would ever read it!”? It sure looks like it.
My guess is that Noziglia, Grimsby and Saunders-Mitchell met in the hallway in the afternoon of 15 August 2016 and agreed for Noziglia to write a phony email that arrives at an amount of dollars aimed to scare me off. In the email below you can read Noziglia suggested to Grimsby to discuss in person in the afternoon of 15 August 2016 the estimate for the costs.
So far we’re confirmed, again, that the US gold is held in secrecy. No surprises there. Moving on to the content of the documents.
Audit Documents Released Are Incomplete
When one walks into a US Mint repository the main barrier will be the door to the vault room. In the case of Fort Knox this a 20-tonne door of which no one person is entrusted with the combination. Once inside the vault room the gold is stored in segregated compartments that are sealed since at least the fifties.
The official narrative is that by 2008 the load of all 42 compartments had been physically audited. Every compartment had been opened, the gold inside counted, weighed and assayed, after which the gold was stacked in an adjacent compartment in the vault room (in several documents it’s described this is the way the gold is physically audited). Subsequently the target compartment door was closed and placed under Official Joint Seal, if during the verification no discrepancies had been found with the Mint’s bullion ledger. In most years until 2008 one or two compartments were opened for a physical bar examination, while the other compartments were merely inspected for any tampering of the Official Joint Seal (OJS). The purpose of joint seals is to avoid the necessity of verifying all assets in each annual audit.
Thus the audits of the Deep Storage gold consist of two conventions: gold verifications, which are the physical audits of gold bars inside the compartments. And OJS inspections, which are checks of the seals placed on the compartment doors. The superintendent in the audit procedures is the Office Inspector General of the US Treasury, in short, the OIG.
When reading the audit documents delivered to me (the Memoranda hereafter) the distinction between gold verifications and OJS inspections is clear. Let me show you an example of Fort Knox. The first screen shots below are from a gold verification at Fort Knox in March 1998.
The following screen shots are from the OJS inspection at Fort Knox in June 1998. We conclude gold verifications and OJS inspections are performed separately and thus are reported as such.
Click here and here to download all Memoranda sent to me by the US Mint.
After I had organized the documents and imported all data in spreadsheets I noted the 134 pages exclude 27 OJS inspection reports and at least 3 gold verification reports. I’ve asked the Mint to deliver the missing Memoranda, although I’m not expecting them to ever comply.
The fact 30 Memoranda are missing is of course highly problematic. Bear in mind, I offered the Mint $3,144.96 dollars to produce these documents.
In case you’re wondering how I know what gold verifications reports I’m missing, this is because references are made to these physical audits in succeeding gold verification reports. Fort OJS inspection reports, those should be done every year.
Below is an example of an Official Joint Seal. I obtained nearly all OJS copies from a separated FOIA request at the OIG.
Submitted a whole bunch of FOIA requests at US government departments regarding Fort Knox today. #gold
Fort Knox Compartment 31 Was Opened In 1996 For Dubious Reasons
There are a couple of disturbing lines written in the Fort Knox OJS inspection report of 1996. Although for an OJS inspection seals should only be examined for tampering, on 12 August 1996 at the Fort Knox OJS inspection two representatives of the General Accounting Office (GAO) showed up in the vault room and decided to select “a single joint sealed compartment for opening and inspection”.
Unfortunately the report doesn’t say what was in the vault compartment; how many bars and fine troy ounces (FTO) it contained. Based purely on this document it would impossible to decipher what the GAO exactly did. However, by combining the info in the 1996 OJS inspection report with documentation obtained through a FOIA requests at the OIG, we do know what happened.
Have another look at exhibit 10. We can read Fort Knox compartment 29 was sealed in 1998. But the content, 19,800 gold bars weighing 6,470,624.049 FTOs before assays samples were taken, was sourced from compartment 31 that was sealed on 12 August 1996. Was compartment 31 the one opened by the GAO in 1996? Yes, without a doubt.
By examining all OJS copies – such as demonstrated in exhibit 10 – it shows there was no other vault segment freshly sealed on 12 August 1996 other than compartment 31. Moreover, the 1996 OJS inspection report mentions only one joint sealed compartment was breached. Therefore we know the GAO representatives opened Fort Knox compartment 31 comprising 19,800 gold bars weighing 6,470,624.049 FTOs on 12 August 1996.
Furthermore, in the 1995 OJS inspection report we read there was one compartment – the number is redacted – that contained 19,800 gold bars weighing 6,470,624.049 FTOs. And in 1995, 1996 and 1997 there were no gold verifications at Fort Knox as far as I know, other than the GAO incident. Have a look below at a screenshot from the 1995 Fort Knox OJS inspection report.
What happened is that on 12 August 1996 compartment 31 was opened by the GAO to “check a few bars”, but then two years later in 1998 the same gold was verified by the OIG; all the gold inside taken out of compartment 31, counted, weighed and assayed, to be stored across the hall in compartment 29. This is suspicious. I quote, “the purpose of joint seals is to avoid the necessity of verifying all assets in each annual audit”.
I do not possess the official rules for US Mint OJS inspection and gold verification for the year 1996 (“MD 8H-1”), but based on the rules that prevailed in 1975, what the GAO did on 12 August 1996 was not done. Read with me.
How come the GAO could open a compartment? The OIG stated under oath in 2011, “since 1993, when we assumed responsibility for the audit, my office has continued to directly observe the inventory and test the gold” (page 4). If the OIG is responsible how come the GAO could break a seal?
Let’s contemplate this: if the “random checks” the GAO performed in 1996 in compartment 31 formed an adequate gold verification, why did the OIG re-audit the exact same gold in 1998? And what was the intention of the GAO in 1996? The GAO couldn’t fully audit compartment 31, because they were present at Fort Knox only for one day (12 August), and no single person or flock of auditors can verify 19,800 large gold bars in one day. The fact these 19,800 gold bars were re-audited in 1998 underlines what the GAO did in 1996 was inappropriate at best.
One theory is that the gold in compartment 31 was prepared in 1996 to be physically audited down the road. Remember what the Fort Knox gold verification report of 1998 stated (exhibit 7.2)? In 1998 the OIG, “selected predetermined individual bars to be drilled for assay”. Possibly, the OIG selected the exact bars in 1998 that were put in in 1996. If this is true the names and autographs of the perpetrators of this crime are on the seal of compartment 29 (exhibit 10).
My succeeding post on this subject will expose that many other Deep Storage compartments at the Mint have been opened for dubious reasons as well. Which could be the reason the Mint didn’t provide us ALL the OJS inspection reports from Denver and West Point from 1993 through 2003 (exhibit 9).
Weighing Sample Size Remarkably Low
We need to discuss the sample size of the gold verifications. In 1998 at Fort Knox 19,800 gold bars were inspected but only 105 of them were weighed and assayed (exhibit 7.2). That’s not much in my humble opinion. In any case, I expected a higher sample size.
In the 1953 audit at Fort Knox (download report here) in total 88,000 bars weighing 48,506,985 FTOs were counted for verification. About 10 % of those were weighed.
During the Continuing Audits from 1974 through 1986 it seems 2 % of the gold counted was weighed. A huge decline from 1953.
Although gold bars tested to be out of tolerance during a Fort Knox audit in 1977 at a sample size of 2 %, by 1998 the sample size had been further debased to 0.53 %. I’m not a professional auditor (if you are one please contact me), but common sense suggests that when irregularities are found the sample size should be increased, not decreased.
To make matters worse, in 1999 at West Point the sample size was 0.52 %, and again, a melt appeared to be out of tolerance.
Was the sample size increased after 1999? Not really. At Fort Knox in July 2000 the samples size was 0.65 % (93 bars weighed of 14,262 bars counted). But wait until I show you what numbnuts were entrusted handling the scale for the audits of the world’s greatest gold hoard.
Scale Didn’t Work, Repeatedly
Let’s study the 2004 physical audit at West Point. Please read:
When all parties tried to reconcile the weight of samples on 22 and 23 July 2004, they found out, “the scale was reading at ounces rather than fine troy ounces”, because, “a setting on the scale had not been properly changed”. Allegedly this is what caused alternative readings in the books of the Director of the Mint’s Representative and the OIG’s Representative. And presumably because nobody could figure out how to use the scale correctly they decided to postpone re-weighing the samples until 24 August 2004. This failure of how to use a scale is a colossal disaster for the credibility of the Deep Storage audit procedures.
In 2004 a mere 71 bars were weighed and assayed, but it appeared that none of the auditors present knew how to rightly use the scale. The Memoranda mentions they found out the scale wasn’t properly functioning when weighing the assay samples, but what about the weighing of the actual bars? What about the weighing of every Deep Storage gold bar under the supervision of the OIG from 1993 until 2008? We have no guarantee this has ever been executed competently.
To repeat, the official explanation for this blunder reads, “the scale was reading at ounces rather than fine troy ounces”, because, “a setting on the scale had not been properly changed”.
First, in my mind there can be no imaginable circumstances in which setting of the scale should have been changed. The scale should read troy ounces to as many decimals all day long. That’s it. Why change the settings?
Second, they say, “the scale was reading at ounces rather than fine troy ounces”, but scales don’t read fine troy ounces so this statement is fake. A scale reads troy ounces, or digital ones can be set to reading grams; it cannot smell what is the purity of the gold and thus display fine troy ounces. That’s what the assay test is for.
In 2008 at West Point a similar disaster happened. Read with me:
The auditors couldn’t clearly read the decimal point. After assay samples were drilled to be taken out, the auditors weighed the same amount of gold granules to replace the samples, in order for the Deep Storage FTOs to remain flat in 2008. But the assay lab, White Sands Missile Range, which is a division of the US Army, found out from the paper work that the weight of the assay samples didn’t match the weight of the granules. And so West Point compartment 10-H had to be re-opened on 22 September 2008 to put an exact 10.346 ounces of gold in, instead of 1.0346 ounces.
What a catastrophe! Be aware that before weighing the granules the auditors weighed 86 gold bars and the assay samples. How do we know they properly weighed the assay samples and the totals of the 86 bars? The short answer is, we don’t.
Thereby, anybody with a sense for gold can see the difference between 10 ounces and 1 ounce of yellow metal.
From the examples above it should be clear that the Deep Storage gold has not been audited by professionals, but the precious metals have been verified by imbeciles. Clearly the scale was repeatedly handled by amateurs, which throws a wrench at the integrity of the entire US official gold reserves auditing project. I’m not at all surprised the US Mint has tried everything to keep the records of the auditors out from the pubic domain. Fortunately most of it will be out in the open eventually. The citizenry of the world deserves to know everything there is about the Deep Storage gold.
Let’s finish with one more comment from the West Point 2006 audit report.
The auditors couldn’t figure how to use the drill to take assay samples (how about pointing the tip to a bar and press the button). They also were oblivious how to calculate fine troy ounces. We must wonder if these people would be capable of tying their own shoelaces. In any case, the fact the US government chose to assign very inexperienced people widely opens the possibility that the audits are a complete hoax.
More will follow…
1. The next list consists of all heros that supported the crowdfunding campaign (mostly Turdites!):
The last bits of data are coming in from the countries that export gold to China, with which we can compute the total the Chinese have imported in 2016. There are four main gold exporters to China, which are Hong Kong, Switzerland, the UK and Australia (it’s not publicly disclosed how much South Africa exports directly to China ). Let’s start discussing the largest gold exporter to China.
Since 2011 when the gold price slowly started to decline and China embarked importing gold at large, Hong Kong has been the main conduit to the mainland. According to data by the Hong Kong Census And Statistics Department (HKCSD) the special administrative region net exported 771 tonnes of gold to China in 2016, ranking first once again. Net exports were down 10 % compared to 2015.
As I mentioned in November 2016 there were rumors that part of the bullion exports from Hong Kong to China were fake – over-invoiced to move capital out of the mainland – which overstated the flow of gold into China. Let’s investigate if the data by the HKCSD can substantiate this rumor. The net amount of bullion going from Hong Kong to China is the residual of exports (materials lastly fabricated in Hong Kong) plus re-exports (materials not altered in any way, shape or form but merely re-distributed by Hong Kong) minus imports (materials imported into Hong Kong from China through processing trade). If one is to engage in over-invoicing exports from Hong Kong are more suitable than re-exports, because the origin of exports are harder to track. For re-exports the origin of the material must be recognized by the HKCSD, which makes any illegal scheme more difficult to conceal.
Notable is that from February through August 2016 there was an increase of gold exports relative to re-exports from Hong Kong to China (see dark green bars in the chart above). Usually the shipments from Hong Kong to China are re-exports, so the increase in exports was remarkable. But the HKCSD data is no hard evidence any transfers were overstated.
In another example: if we look at the composition of Hong Kong’s export and re-export to the UK in 2016, we can see something similar, the majority were exports.
I doubt Hong Kong’s flow of gold to the UK has been overstated; UK residents have no motive to surreptitiously move capital abroad. And if the data on Hong Kong’s shipments to the UK are accurate, why can’t the data on Hong Kong’s shipments to China be accurate? Thereby, the Chinese customs department is not retarded. I’m quite sure the Chinese customs department is aware of over-invoicing schemes and as a consequence it can strictly monitor cross-border gold flows. My conclusion is that net shipments from Hong Kong to China in 2016 have likely been close to 771 tonnes. If I do ever find hard evidence it was less I will report accordingly.
Most likely Hong Kong’s position as the largest gold exporter to China will slowly fade in the coming years, as the State Council is stimulating gold freight to go directly to Chinese cities (hoping the Shanghai International Gold Exchange will eventually overtake Hong Kong’s role as the primary gold hub in the region). Consequently, gold exports to China are increasingly bypassing Hong Kong.
In December 2016 we got a preview of what is about to come: Switzerland net exported an astonishing 158 tonnes directly to China, up 418 % from November 2016, up 168 % from December 2015, and 106 tonnes more than what Hong Kong did.
It will take more time before Hong Kong’s role as supplier to China is fully over though. In the past years a significant part of gold exports to China has been used to quench Chinese jewelry demand. The core of the Chinese jewelry manufacturing industry is located in Shenzhen, which is right across the border from Hong Kong. But as far as I know there aren’t many flights going directly from Switzerland, Australia or the UK to Shenzhen yet. Hong Kong on the other is well connected; so in the near future Hong Kong’s airport is more convenient to supply gold from abroad to Chinese jewelry manufacturers.
In total the Swiss net exported 442 tonnes directly to China mainland in 2016, up 53 % from 288 tonnes in 2015.
The United Kingdom
Direct gold shipments from the UK to China have been tepid in 2016. Only 15 tonnes have been net exported to the mainland over this time horizon. Noteworthy though, in December 2016 the UK exported 172 tonnes to Switzerland, which in turn moved 158 tonnes to China – as I mentioned in the previous chapter. So although the UK didn’t directly export metal to China in December, it sure was the main supplier.
Not all data from the Australian Bureau of Statistics (ABS) has been released for 2016, but from what we have Australia seems to have exported less to China than in 2015. From January through September direct shipments amalgamated to 53 tonnes, while Australia directly net exported 78 tonnes to China over the same months in 2015. (ABS has notified me they changed the way how they disclose gold export data, but they failed to clarify the details. Until I get more information I will stick to my own formula to compute Australia’s direct export to China, which was confirmed to be accurate by ABS early 2016.)
It can be Australia’s direct exports where strong in last three months of 2016 as the price of gold went down over this period and the Chinese increase gold purchases on a declining gold price.
Combining gold trade data by Hong Kong, Switzerland, the UK and Australia, reveals China has imported at least 1,281 tonnes in 2016. Though this figure excludes Australia’s exports for October, November and December, so I’m estimating total Chinese gold import will reach roughly 1,300 tonnes.
According to the data at my disposal there have been practically zero tonnes of gold imported from China by other nations across the globe than the ones discussed above. Signalling there is very little gold being exported from the Shanghai International Gold Exchange (SGEI) located in the Shanghai Free Trade Zone. Possibly foreign central banks buy gold on the SGEI and ship it home as monetary gold which doesn’t show up in any customs reports. However, in the history of the SGE/SGEI a mere 3 tonnes has been traded in 12.5 Kg bars, all the rest was in smaller bars, mainly 1 Kg. And I assume central banks would prefer large bars. All in all I think that up till now the SGEI has mainly been used by Chinese banks to import gold from the Shanghai Free Trade Zone into the domestic market.
A few weeks ago I estimated Chinese gold import 2016 would aggregate to 1,300 tonnes, to which I calculated 5,000 tonnes of gold have been moved into the Chinese domestic market from 2007 through 2016 on top op the imports to satisfy Chinese consumer demand. In my post The West Has Been Selling Gold Into A Black Hole I explain how I think this will strengthen a forthcoming gold bull market.
Last but not least: SGE withdrawals for January 2017 came in at 184 tonnes, down 18 % from January 2016.
This post is part of the Chinese Gold Market Essentials series. Click here to go to an overview of all Chinese Gold Market Essentials for a comprehensive understanding of the largest physical gold market globally. This post was updated in 2017.
The difference between SGE withdrawals and Chinese consumer gold demand as disclosed by GFMS has aggregated to a staggering 6,032 tonnes from 2007 until 2016 (the period this article will focus on). To explain the difference, GFMS and other Western consultancy firms have presented several arguments in publications and lectures at conferences throughout the years, though none of them can really explain the difference in full. This post is an overview of all such arguments – supplemented by my own arguments.
The reason I tend to compare SGE withdrawals to Chinese gold demand as disclosed by GFMS, and not Metals Focus or CPM Group, is because the GFMS is globally the easiest (free) accessible data source for investors (next to the World Gold Council). Usually investors and news agencies worldwide consult GFMS (or WGC) for supply and demand statistics, which make this the most important firm to test for accurate numbers. Below we’ll examine to what degree the arguments can or cannot have caused the difference.
This is the argument list (by GFMS, WGC and CPM Group) in chronological order:
Industrial demand (August 2013)
Stock movement change (August 2013)
Round tripping (April 2014)
Leasing (April 2014)
Official purchases (April 2014)
Recycled distortion (November 2014, February 2015, March 2016)
Export (May 2015)
Chinese commercial banks’ balance sheets
Financial statement window dressing (March 2016)
Retailers selling unsold inventories directly to refiners (March 2016)
Surprisingly, after the reports were published and I had debunked the arguments in it, the WGC and GFMS swiftly came up with brand new arguments. Note this shift in arguments; when the old ones failed, the firms impudently moved on and came up with new ones. The fact the argument list is constantly changing confirms the weakness of all arguments it holds, and the apparent ‘ignorance’ of Western consultancy firms regarding the Chinese gold market. (By the way, eventually in late 2016 I found out why Western consultancy firms lie about Chinese gold demand, which is explained in my post The Great Physical Gold Supply & Demand Illusion.)
First, let’s go through all the arguments to investigate which ones make any sense. At the end of the post we’ll do some number crunching.
1) INDUSTRIAL DEMAND. The first argument ever presented to me came from the WGC. In August 2013 I’ve asked the Council what their explanation was for the difference between their Chinese gold demand numbers and demand as disclosed in the CGA Gold Yearbooks co-written by the PBOC – the latter exactly equals SGE withdrawals. They replied to me by email:
The data that we publish in Gold Demand Trends are collected for us by Thomson Reuters GFMS. Our data represent jewelry and bar & coin demand and do not incorporate any industrial demand or fabrication, which is included in the PBoC figures. As I am sure you will appreciate, data collection of this sort relies on a number of proprietary sources and these will not necessarily be the same for both GFMS and PBOC. It is, therefore, perhaps not surprising that the estimates of demand differ somewhat.
The WGC identified a gap of, at that point, 2,000 tonnes of gold with industrial demand. Not very credible. Additionally, in Q1 2015 the WGC started including industrial demand (technology) in its data, and as far as I know GFMS has always included industrial demand in its data. So, for our comparison of SGE withdrawals versus GFMS demand this argument is irrelevant.
2) STOCK MOVEMENT CHANGE. When I asked GFMS in August 2013 about net investment – which is how the difference was titled in the CGA Gold Yearbooks – they wrote me by email:
We have checked with our Data Specialist and confirmed that we use a different methodology. Total Chinese demand used by Thomson Reuters GFMS only includes jewelry, physical bullion bars/coins and all industrial demand. Any stock movement change (which is essentially the item 6 net investment) will not be included as underlying demand.
So according to you category six is “stock movement change”? This would be gold added to the stocks from jewelers, the mint, industrial companies, etc? (this is a few hundred tons each year!)
That’s correct based on the resolution provided by our data specialist.
Because SGE withdrawals capture wholesale demand the difference is partially what jewelry companies, refineries, industrial companies and the mint have purchased at the SGE, but not yet sold in retail. And so, stock movement change is a legitimate argument, though the amount of gold in stock can never explain the full difference of 6,032 tonnes.
According to an estimate by the WGC as much as 125 tonnes of gold can have been absorbed as inventory in the Chinese domestic gold market from 2009 until 2013:
… It is, however, indicative that as jewelers expanded, so too did their inventory levels and it is our judgment that across the industry between 75t to 125t may have been absorbed in the supply chain since 2009.
Stock movement change is a legitimate argument and its volume, 125 tonnes, will be taken into account for our calculation of true Chinese gold demand at the end of this post. I will stick to the number 125 because in my opinion the jewelry and coin industry in China hasn’t grown since 2013 (meaning inventory stayed flat).
3) ROUND TRIPPING. In April 2014 the WGC published a report titled China’s Gold Market: Progress and Prospects. It certainly was not the first WGC report on China – in 2010 China Gold Report was released – but it was the first time the Council elaborated on the structure of the Chinese gold market, the Shanghai Gold Exchange and the “supply surplus” in the Chinese gold market. Logically, the Council had some explaining to do, as it was clear China imported substantially more gold than what they disclosed as demand.
For the first time Chinese Commodity Financing Deals (CCFD) were introduced to the Council’s wide reader base. This type of financing is pursued to acquire cheap funds. It can be done trough round tripping or gold leasing. The Council wrote:
These operations fall into two broad categories, although there is some overlap between the two. Firstly, there is the use of gold via loans and through letters of credit (LCs) as a form of financing. Secondly, there is the use of gold for financial arbitrage operations that will also be based upon gold loans or LCs. In most cases the gold is quickly re-exported to Hong Kong, often as very crude jewellery or ornaments to get round tight controls on bullion exports. (This is the practise commonly referred to as ‘round-tripping’. Moreover, because nearly all gold flowing into China goes through the SGE, round-tripping can inflate the SGE delivery figures.) In other cases the metal is stockpiled in vaults in China or Hong Kong.
So round tripping is not a legitimate argument. To my understanding the WGC has abandoned this argument all together, though GFMS still thinks round tripping inflates SGE withdrawals. In their Gold Survey 2015 it’s written (page 78):
…the round tripping flows between Hong Kong and the Chinese mainland, which also inflates the SGE turnover and withdrawal figures…
4) GOLD LEASING. The other CCFD is leasing. In the WGC report from April 2014 it’s stated:
No statistics are available on the outstanding amount of gold tied up in financial operations linked to shadow banking but Precious Metals Insights [PMI] believes it is feasible that by the end of 2013 this could have reached a cumulative 1,000t…
5) OFFICIAL PURCHASES. Often it’s being thought in the gold space SGE withdrawals end up in the vaults from the People’s Bank Of China (PBOC). Early 2014 the WGC (in China’s Gold Market: Progress And Prospects) speculated the difference could be explained by official purchases, though, later that year the Council changed its mind. From the July 2014 WGC report on China, Understanding China’s Gold Market, we can read:
China’s authorities have a range of options when purchasing gold. They may acquire some of the gold which flows into China; there has been no shortage of that. But there are reasons why they may prefer to buy gold on international markets: gold sold on the SGE is priced in yuan and prospective buyers – for example, the PBoC with large multi-currency reserves – may rather use US dollars than purchasing domestically-priced gold. The international market would have a lot more liquidity too.
6) RECYCLED DISTORTION. The most obvious argument to explain elevated SGE withdrawals, one would think, is recycled gold through the bourse counted over and over as withdrawn. However, SGE rules state bars withdrawn are not permitted to re-enter the vaults before being remelted and assayed by an SGE approved refinery. Which is not say it doesn’t happen.
Arguments presented by the firms regarding recycled gold must be divided in subcategories. There is process scrap, arbitrage refining, and there are VAT schemes.
6.1) Process scrap. This argument was first presented by CPM Group. In short, CPM states industrial companies produce 50 – 70 % scrap supply of the gold used in manufacturing. The scrap spillover flows directly back to the SGE. Process scrap thus inflates SGE demand and supply, because the gold was bought at the SGE (demand), but a significant part flows back to the SGE (supply). The part that is recycled through the SGE has no impact on the price.
Although, it’s unknown how much of process scrap actually flows back to the SGE or is brought to a refinery for toll refining (a refinery producing bars or wire from the process scrap for the industrial company in return for a fee).
Process scrap, described first in detail by Jeffrey Christian in November 2014 in the chapter “CPM Group” at the very end of this post, is a form of recycled distortion, and is a legitimate argument.
6.2) Arbitrage refining. This argument was brought forward by GFMS on 17 February 2015 at the Reuters Global Gold Forum when Jan Harvey interviewed Samson Li (GFMS).
Some people see withdrawals on the Shanghai Gold Exchange as a proxy for Chinese demand. Do you think this is valid?
It depends on the methodology used. For example there are refiners that would, at times, withdraw 9995 gold bars from the SGE, refine it into 9999 bars whenever there is profitable opportunity, and then deposit it back into SGE vault……
Presumably, there can be an arbitrage opportunity at the SGE if Au99.95 gold is an X percentage cheaper than Au99.99 gold. Such a spread would be a classic example of one of the contracts being under or overvalued relative to the other.
I’m not a trader, but I can imagine a way to close the arbitrage through gold leasing. This is my theory: if a spread occurs Au99.95 is bought, concurrently Au99.99 (LAu99.99) is borrowed and immediately sold. Then the Au99.95 is withdrawn, refined into Au99.99 and returned to the lender.
If the arbitrage described above exists, inter alia depends on the speed to which a lease contract can be settled. If a spread occurs and the refiner has to wait 2 days before it can take delivery of Au9999, the arbitrage won’t fly. I’ve asked the ICBC gold lease desk what would be the fastest possibility to sign a lease contract. They told me usually it takes several days or weeks as the lessee’s credit rating must be determined. Though, for regular customers the lease ca be executed in one hour.
It’s hard for me to say if arbitrage refining is really possible according to the aforementioned theory, because it depends on many variables and the established relationship between lessor and lessee. In addition, why would anybody sell Au99.95 if it was undervalued? In my opinion the argument that arbitrage refining inflates SGE withdrawals can be doubted.
6.3) VAT schemes. This argument brought forward by GFMS in The Gold Survey 2016 is legitimate. Though, it’s unknown to what extent it has been used. Read more about thevalue-added tax system in China’s domestic gold market by clicking here, and read why I think the VAT scheme can only have had a limited impact on SGE withdrawals in the chapter “Tax Avoidance” inthis post.
7) EXPORT. This argument was brought forward by PMI. On a conference in London (2 May 2015) Phillip Klapwijk, Managing Director of Precious Metals Insights Limited (PMI), stated China exports about 1,000 tonnes a year from the domestic gold market. However, at this stage the rules prohibit gold export from the Chinese domestic gold market. I’ve written an extensive analysis on Klapwijk’s presentation (click to read), no need to go over this again here. The export argument is not legitimate.
8) CHINESE COMMERCIAL BANK BALANCE SHEETS. Over the years, on countless gold blogs the “precious metals” (more than 2,500 tonnes by now) on Chinese commercial banks’ balance sheets have been identified as the “surplus” in the Chinese gold market. But not according to my research. After a thorough study I think the gold on the banks’ balance sheets reflect a mixture of GAP gold, retail inventory, gold held for hedging, gold outside China, but most importantly back-to-back leasing and synthetic leasing. The leasing business by banks can make it appear the banks “own” gold assets, while in fact it’s just accounting that makes it seem that way.
9) FINANCIAL STATEMENT WINDOW DRESSING. Another argument that was presented by GFMS in their Gold Survey 2016. In short, the argument is false. If you want to know why please read the chapter “Financial Statement Window Dressing” in this post.
10) RETAILER SELLING UNSOLD INVENTORIES DIRECTLY TO REFINERS. Another argument that was presented by GFMS in their Gold Survey 2016. The argument can be true. If you want to learn more please read the chapter “Retailers Selling Unsold Inventories Directly to Refiners” in this post.
11) THE SHANGHAI INTERNATIONAL GOLD EXCHANGE. This argument was conceived by myself. As we could have read in The Workings Of The Shanghai International Gold Exchange(andSGE Withdrawals In Perspective), the gold withdrawn from the SGEI vaults in the Shanghai Free Trade Zone (SFTZ) – note, this tonnage is included in “SGE withdrawals” – can, (i) either be imported into the Chinese domestic gold market, or, (ii) exported abroad and thereby distorting Chinese wholesale gold demand when measured by SGE withdrawals. However, up until December 2015 we know SGEI withdrawals have rarely been exported abroad, according to several sources.
Another way of checking possible SGEI withdrawals that have been exported abroad is simply examining the gold imports of all countries on earth from China since 2014. As far as I can see the exports from China have been tepid. For example, 7 tonnes were exported to Thailand in 2015 and 8 tonnes to the UK in 2016, but that’s about it. Except for Chinese exports to Hong Kong: in 2016 these accounted for roughly 90 tonnes. But, most likely this gold didn’t come from the SFTZ (SGEI), but from the Shenzhen Free Trade Zone just across the border from Hong Kong where China’s jewelry manufacturing base is located. Gold export from the Shenzhen Free Trade Zone to Hong Kong has existed long before the SGEI was erected (and these flows are offset when I compute net gold flows into China from Hong Kong). How it works is that Shenzhen gold manufacturers import gold through processing trade, fabricate the materials into jewelry and ornaments after which the finished products are exported through processing trade back to Hong Kong where they are sold locally or distributed across Asia.
All in all, I’m still not seeing a lot of gold being withdrawn from SGEI vaults and exported abroad.
One last possibility that would inflate “SGE withdrawals” is when (Asian) central banks buy gold on the SGEI, after which it’s withdrawn and exported. Because central banks can monetise gold, which is exempt from being disclosed in customs data, we would never see these exports out of the SFTZ. Effectively we would only see export to China and high SGE withdrawals, while gold is withdrawn in the SFTZ and covertly exported abroad. At this point not a very plausible scenario because SGEI gold trades at a premium from gold in London, Singapore, Hong Kong, etc, but certainly possible.
12) SMUGGLING. Naturally, smuggling can cause SGE withdrawals to be inflated. Indians could buy gold in Shenzhen, withdraw from the SGE vault and smuggle it home. Although, we have no numbers on smuggling so I can’t take it into account for our calculation of true Chinese gold demand.
Unfortunately we don’t know how much disinvestment is in China, and as a consequence neither do we know recycled distortion. But we’ll set the lower bound by apparent supply and upper bound by SGE withdrawals, and work from there.
From 2007 until 2016 apparent supply in the Chinese domestic gold market – assuming all net exports from the UK, Switzerland, Australia and Hong Kong went to the domestic market and SGEI vault inventory is insignificant – was 12,183 tonnes, and total SGE withdrawals over this period accounted for 13,388 tonnes.
So the lower bound of true Chinese gold demand over 2007-2016 was 12,183 tonnes, and the upper bound was 13,388 tonnes. Adjusting the upper bound by wholesale inventory increase (125 tonnes) makes 13,263 tonnes. Effectively, true Chinese gold demand must have been somewhere in between 12,183tonnes and 13,263 tonnes (instead of the 7,356 tonnes GFMS has presented the world). Let’s, for the sake of simplicity, take the middle of the 12,183 and 13,263 as a number for true Chinese gold demand to work with – which is 12,723 tonnes.
The gap between GFMS demand (7,356 tonnes) and our estimate of true demand (12,723 tonnes) is a staggering 5,367 tonnes. It’s impossible to deny this immense tonnage has not been demand from high net worth individuals and institutions, purchased directly at the SGE. Although, the remaining difference between 12,723 tonnes and 13,263 tonnes could have been caused by process scrap, VAT schemes, retailers selling unsold inventory to refiners etc.
Ironically, the WGC wrote in its recent GDT Q2 2017:
Purchases made directly from the SGE continued to gain traction accounting for a significant proportion of Q2 bar demand. Investors benefit from better pricing bars from the SGE are usually 5-10 yuan per gram lower than those bought from commercial banks – and a sense of security from knowing they are buying gold from a trusted provider. With a minimum lot size of 100 gram direct withdrawals from the SGE largely serve China’s high net worth individuals.
Finally, after four years of debating (between me and the WGC/GFMS) the WGC admits that the majority of the difference simply reflects direct purchases of high net worth individuals and institutions at the SGE. Case closed.
To this date, October 2017, I still think SGE withdrawals provide a useful indicator of Chinese wholesale gold demand.
In December 2016 Chinese wholesale gold demand, measured by withdrawals from the vaults of the Shanghai Gold Exchange (SGE), accounted for 196 tonnes, down 9 % from November. December was still a strong month for SGE withdrawals due to the fact the gold price trended lower before briefly spiking at the end of the month, and the Chinese prefer to buy gold when the price declines (see exhibit 1).
In total Chinese wholesale gold demand reached an astonishing 1,970 tonnes in 2016. But will these huge tonnages bought by China ever have an impact on the gold price? I think it will.
As in previous years, SGE withdrawals were mostly supplied through imports, in 2016 at approximately 1,300 tonnes. And as in previous years, SGE withdrawals were roughly twice the size of Chinese consumer gold demand. The latter is published by all “leading” consultancy firms, such as the World Gold Council and Thomson Reuters GFMS. Because these firms have systematically underreported and eclipsed Chinese gold demand since 2007, a significant share of the financial industry is unaware China has imported 5,000 tonnes in the past years, which is not allowed to be exported. My hypothesis is that this 5,000 tonnes decline in above ground gold reserves outside of the Chinese domestic market will make gold rally stronger in a future bull market than it did in previous bull markets. To the extent many investors are uninformed about the shrinking volume of troy ounces available outside of China, their ignorance will boost any price rally coming.
In this post I would like to share my thoughts on how the gold price is correlated to trade in above ground reserves, and how China has slashed these reserves to the tune of 5,000 tonnes, which will significantly impact the next leg up in gold.
Correlated: The Gold Price And UK Gold Trade
Since many decades large investors in the West set the price of gold. Ever since, the heart of the Western gold wholesale market has been London in the United Kingdom. There is thus a correlation between the gold price and the volume of gold net imported or exported by the UK.
In Asia, on the other hand, gold market participants are more price sensitive, implying they buy low and sell high (the opposite of Western investors). I’ve described this trend frequently on these pages, but the same can be read in books by gold author Timothy Green. In The Prospect For Gold from 1987 Green states:
Before we discuss the connection between Western supply and demand trends to developments in the Chinese gold market of the past decade, let me first recapitulate that global physical gold supply and demand is far in excess of the statistics the World Gold Council and GFMS publish. Below is a chart that shows the quarterly averages of all physical supply and demand categories as disclosed by the World Gold Council from Q1 2002 until Q4 2015. These numbers are more or less the same as figures by GFMS.
We can see that over the course of 13 years, the majority of supply consisted of mine output (73%) and the majority of demand consisted of jewelry consumption (64%).
(Note, the categories official sector, net producer hedging and ETFs can be either supply or demand and volumes can greatly vary per quarter. Though, only in 1 of 52 quarters examined has ETF demand been greater than jewelry consumption (Q1 2009). In all other quarters official sector, net producer hedging and ETFs supply or demand has not been greater than mine output or jewelry consumption.)
If the data by the World Gold Council regarding physical gold supply and demand would be exhaustive, mine output and jewelry consumption should have a positive correlation to each other and the price of gold. But they don’t. Have a look at the next chart.
During the bull market from 2002 until 2011 jewelry consumption decreased and it hardly ever transcended mine output. In turn, mine output gradually ascended over this time horizon while the gold price increased six fold! Are the forces between jewelry demand and mine supply driving the medium/long term price of gold? No, clearly not. This shows the data by the World Gold Council is incomplete.
(I should add that mine output does have a correlation to the gold price in the very long term as it can take more than ten years to setup a gold mining project. See the next chart.)
In contrast to the data by the World Gold Council, we can observe a strong correlation between the medium/long term gold price and institutional supply and demand flowing through London. View the chart below.
Strangely, institutional supply and demand are categories not included in the World Gold Council’s data – or in any other precious metals consultancy firm’s data that I’m aware of.
Because in the UK there are no refineries, no gold mines and local consumption demand and scrap supply is immaterial, all gold that is visibly (non-monetary) imported and exported must either relate to ETF holdings stored in London, or Western institutional supply and demand. When we compute the ratio between both, ETF flows compound to roughly 35 % of the UK’s net flow (import minus export) and as a consequence approximately 65 % is Western institutional supply and demand. Effectively the majority of the UK’s net flow is Western institutional supply and demand.
Hereby, consider that all supply and demand categories disclosed by the World Gold Council more or less equal each other (exhibit 4), so for the sake of simplicity we‘ll state that total mine output + scrap supply versus jewelry consumption + bar and coin + industrial demand meets outside the UK and doesn’t set the medium/long term price of gold.
The UK’s net flow, on the other hand, is highly correlated to the medium/long term price of gold. Note how nearly every month the change in net flow corresponds with the direction of the gold price (exhibit 6). Less granular, from the moment my data starts in 2005 the UK has been a net importer until 2012 on a rising price of gold. From 2013 until 2015 the UK was a net exporter on a declining price of gold. And in the first quarter of 2016, when the gold price saw its strongest move up since 1986, the UK was a net importer. Coincidence? I think not.
We can conclude that Western institutional supply and demand in above ground gold reserves is driving the medium/long term price of gold. As it’s likely the price of gold could not have gone up from 2002 until 2011 if there had been no UK net imports, and it’s likely the price of gold could not have gone down from 2013 until 2015 if there had been no UK net exports. (Short term the gold price is pushed around in the paper markets.)
We can think of Western institutional supply and demand (the UK net flow) like this: the majority of the gold gross imported into the UK is demand from above ground reserves outside the UK, and the majority of the gold gross exported from the UK is supply to above ground reserves outside the UK. When the UK is a net importer that means there is a net pull on above ground reserves outside the UK, which corresponds to a rising gold price. When the UK is a net exporter the inverse is true.
Here it becomes apparent that the amount of above ground bullion is essential for future price developments.
The Chinese Black Hole
Let’s turn to China. In the introduction I stated China is importing a lot more gold than is known in the financial industry because most investors base their knowledge on data by the World Gold Council. More precise, China has imported 5,000 tonnes from 2007 until 2016 in addition to what the World Gold Council has portrayed through their demand statistics.
Let’s get our minds around this through some charts. As an example, I’ve drawn a chart showing Chinese gold supply and demand for 2015 (last year I have complete data of).
We don’t know every exact data point for China, but we do know GFMSdemand (purple) and apparent supply, consisting of domestic mine output (green), scrap supply (yellow) and net import (blue). From here on we’ll use GFMS data, as GFMS publishes scrap supply numbers for China and the World Gold Council doesn’t.
According to GFMS Chinese consumer gold demand in 2015 was 867 tonnes. To meet demand GFMS presents 450 tonnes was domestically mined and scrap supply accounted for 225 tonnes. Indirectly GFMS states China net imported 192 tonnes to complete the supply and demand balance in the Chinese domestic market (exhibit 7). For the additional 1,383 tonnes imported GFMS has floated all sorts of excuses, which I‘ve debunked here and here.
The bottom line is, in addition to the 192 tonnes GFMS reports as imported in 2015 to meet consumer demand, China imported 1,383 tonnes to meet institutional demand and all this metal is not allowed to be exported.
If we repeat the same exercise for every years since 2007, the aggregated net imports by China that have not been included in the statistics by GFMS account for 5,000 tonnes. See the next chart.
You can see now, China has enormously diminished above ground reserves outside of the Chinese domestic market without all investors around the world being fully aware. In my humble opinion this will make the price of gold go up turbo charged next time the West shows interest in the metal.
In The Prospect For Gold Green states:
“Selling gold is not a one way street”, wrote Green in 1987. But guess what. Since a few years – from the moment China became an elephant player in the physical market – selling gold is a one way street! Western sell-offs are transhipped to China but do not return. The global gold game has changed.
The consequence is that there are less above ground reserves outside of China for Western investors to buy in a forthcoming bull market, which will elevate the dollar bid per unit gold – in other words the gold price measured in US dollars per troy ounce.
Keep in mind, this phenomenon (China importing vast quantities in addition to Chinese consumer gold demand as disclosed by GFMS) has greatly materialized in 2013, when gold entered a bear market after an 11-year run up. In the previous bull market (2002-2012) above ground reserves outside of China had not been slashed yet. So the ramifications of this phenomenon will only be felt during the next leg up.
Is there any proof to substantiate my hypothesis? I think so. Early 2016 there was some renewed interest in yellow metal from large Western investors. When the price of gold started to climb it went practically vertical ending the first quarter of 2016 up 16.7 %, the strongest quarter since 1986. Coincidence? I think not. It went up strong as it did because there were fewer ounces in above ground reserves available.
A study on how much above ground reserves there are outside China will be saved for a future blog post.
This post is part of the Chinese Gold Market Essentials series. Click here to go to an overview of all Chinese Gold Market Essentials for a comprehensive understanding of the largest physical gold market globally. This post was updated late 2017.
The main arguments presented by Western consultancy firms, such as GFMS and the World Gold Council (WGC), to explain the difference between SGE withdrawals and Chinese consumer gold demand relate to Chinese Commodity Financing Deals (CCFDs). However, this analysis is incorrect as I will demonstrate in this post.
CCFDs are used by Chinese speculators to acquire cheap funds using commodities as collateral. When it comes to using gold as collateral for CCFDs there are two options: round tripping and gold leasing. First we’ll discuss round tripping.
Goldman Sachs (GS) has properly described the round tripping process in a report dated March 2014. We’ll start by reading a few segments from GS about financing deals [brackets added by me]:
While commodity financing [round tripping] deals are very complicated, the general idea is that arbitrageurs borrow short-term FX loans from onshore banks in the form of LC (letter of credit) to import commodities and then re-export the warrants (a document issued by logistic companies which represent the ownership of the underlying asset) to bring in the low cost foreign capital (hot money) and then circulate the whole process several times per year. As a result, the total outstanding FX loans associated with these commodity financing deals is determined by:
– the volume of physical inventories that is involved
– commodity prices
– the number of circulations
Our understanding is that the commodities that are involved in the financing deals include gold, copper, iron ore, and to a lesser extent, nickel, zinc, aluminum, soybean, palm oil and rubber.
…Chinese gold financing deals are processed in a different way compared with copper financing deals, though both are aimed at facilitating low cost foreign capital inflow to China. Specifically, gold financing deals involve the physical import of gold and export of gold semi-fabricated products to bring the FX into China; as a result, China’s trade data does reflect, at least partially, the scale of China gold financing deals. In contrast, Chinese copper financing deals do not need to physically move the physical copper in and out of China, so it is not shown in trade data published by China customs. In detail, Chinese gold financing deals includes four steps:
Onshore gold manufacturers pay LCs to offshore subsidiaries and import gold from Hong Kong to mainland China – inflating import numbers
offshore subsidiaries borrow USD from offshore banks via collaterizing LCs received
onshore manufacturers get paid by USD from offshore subsidiaries and export the gold semi-fabricated products – inflating export numbers
repeat step 1-3
Important to understand is that gold in round tripping needs to be physically imported into China and then exported, in contrast to copper. The reason for this, which GS fails to mention, is that the cross-border trade rules for gold in China are different than for all other commodities. Only through processing trade gold can be imported into China mainland by enterprises that do not carry a PBOC gold trade license. Round tripping by speculators can only be done through processing trade, as it’s not possible through general trade to ship gold into China without a PBOC license. Consequently, round tripping flows are completely separated from the Chinese domestic gold market where the SGE operates. And hence, round tripping cannot inflate SGE withdrawals.
Only by bending the rules – set up a fake jewelry enterprise in a CSSA – speculators can import gold to round trip. By using processing trade, in order to import gold into China, speculators are required to subsequently export the exact same amount of gold, because these speculators pretend to be jewelry manufacturers importing gold for genuine production, which upon completion must be exported. This is why the gold is round tripped. The requirement for export in processing trade can be read in the official PRC Customs Supervision and Administration of Processing Trade Goods Procedures (2004):
“Processing trade” shall refer to the business activity of import of operating enterprises of all or some raw and auxiliary materials, components, parts, mechanical components and packing materials (Materials and Parts) and the re-export thereof as finished products after processing or assembling.
Now we can understand why GS wrote [brackets added by me]:
Specifically, gold financing deals [round tripping] involve the physical import of gold and export of gold semi-fabricated products to bring the FX into China…
The speculators export semi-fabricated gold products to keep up the appearance they are genuine gold fabricators, for which the gold imported must be processed and exported.
On a side note, the amount of gold used in round tripping can be at most the amount of gold yearly exported from China (to Hong Kong). Though the total exported gold will also contain genuine processing trade, so round tripping will likely be less than this amount.
Round tripping does not inflate net export from Hong Kong to China, only gross trade. The net amount of gold imported into China is shipped through general trade, via the SGE, into the Chinese domestic gold market and is prohibited from being exported.
In the chart above we can see China exported 330 tonnes to Hong Kong in 2013. Let’s guess 200 tonnes of that was genuine processing trade (jewelry manufactured in a Chinese CSSA).
330 – 200 = 130 tonnes
Possibly, there was 130 tonnes imported into China for round tripping and subsequently exported back to Hong Kong. Or, 10 tonnes was imported into China for round tripping and subsequently exported to be round tripped an additional 12 cycles, making 13 rounds in total.
13 x 10 = 130 tonnes
In the latter scenario a lot less physical gold is involved (10 tonnes versus 130 tonnes). In reality it’s more likely a gold batch used in round tripping is making multiple rounds than one round.
The Chinese Gold Lease Market
The other gold financing deal that can be conducted by Chinese speculators is gold leasing (which is the same as gold lending). In general gold leasing is a normal market practice.
To become familiar with the lease market I have categorized all potential gold lessees (borrowers) in three groups for us to have a look at examples (with US dollars) of how gold lending is done in financial markets:
A gold miner needs funds to invest in new production goods. It can borrow dollars from a bank at a 7 % interest rate, or borrow gold at 2 % – the gold lease rate is usuallylower than the dollar interest rate. The miner chooses to borrow 10,000 ounces and sells it spot at $1,500 an ounce. The proceeds are $15,000,000 that can be used to invest in new production goods. In a years time the miner has mined 10,200 ounces to repay the principal debt plus interest (the interest on gold loans can be settled in gold or dollars, depending on the contract). Through gold leasing the miner has acquired cheap funding compared to a dollar loan.
A jeweler needs funds to buy gold stock for production. It can borrow dollars from a bank for 7 %, or borrow gold for 2 %. The jeweler borrows 10,000 ounces of gold, with which it can start fabricating jewelry. To hedge itself against price fluctuations the jeweler can sell spot, for example, 10 % of the 10,000 ounces it has borrowed (1,000 ounces at $1,500 makes $1,500,000) to buy gold futures contracts in order to lock in a future price. After a year the jeweler has sold the 9,000 ounces (as jewelry) for dollars and can take delivery of the long futures contracts to repay the gold loan.
A speculator is looking for cheap funds. It can borrow dollars from a bank for 7 %, or borrow gold for 2 %. He borrows 10,000 ounces and sells it spot at $1,500 an ounce. The proceeds are $15,000,000 and subsequently these newly acquired funds can be used to invest in higher yielding products (> 2 %). If the trader chooses to hedge itself in the futures market is up to him. After a year the 10,000 ounces plus interest need to be repaid, either the trader can purchase gold with the profits made on the higher yielding investment or from delivery of futures contracts.
…the SGE provides a crucial role in gold leasing. The SGE’s block trading system is the trading platform used by gold leasing participants; the SGE also provides transfer and settlement services.
China’s gold leasing does not involve the central bank. Gold leasing takes place between commercial banks and enterprises as well as between commercial banks, the former being key.…
An enterprise that intends to be a lessee approaches a branch office of a commercial bank with a rate request and application.
The commercial bank carries out due diligence and then submits a review to their head office for approval.
Upon approval the head office quotes a lease rate with the international gold lease rate as a benchmark plus additional basis points taking into account the potential lessee’s credit, physical gold management costs and other factors.
If the potential lessee accepts the offer, a commercial bank branch manager will sign a lease contract with the customer including the terms and conditions clearly laid out.
According to the “Shanghai Gold Exchange Lease Transfer Procedure”, after signing the lease, the head office of the commercial bank and lessee, or his agent, shall make a lease application through the exchange’s membership system. After verification, the SGE shall transfer the commercial bank’s gold from its SGE bullion account to the lessee’s SGE bullion account. The lessee can now trade the physical gold that it has leased or withdrawal the gold from the vaults.
Upon expiration of the lease the lessee shall deposit or purchase physical gold through the SGE to repay the gold. Corresponding physical gold will be transferred from the lessee’s SGE bullion account to the commercial bank’s bullion account. Leasing fees involved will be settled in currency. At this point, the lease is completed.
I would like to insert a comment supplementing the PBOC’s description of gold leasing in the Chinese domestic gold market. In the paper it says:
“After verification, the SGE shall transfer the commercial bank’s gold from its SGE bullion account to the lessee’s SGE bullion account. The lessee can now trade the physical gold that it has leased or withdrawal the gold from vaults.”
My source at ICBC’s precious metals trading desk told me ICBC has little gold of itself for leasing, most of the gold lend out is sourced from third parties. These parties are either SGE members or overseas banks that supply gold through the Chinese OTC market. ICBC operates in the lease market as an intermediary by connecting supply and demand, it can lease from international banks or local gold owners with an SGE Bullion Account and lend the gold to miners, jewelers or speculators. My suspicion is that the international gold lease rate is lower than the Chinese gold lease rate, which can attract gold from the international market into the Chinese domestic gold market.
The Chinese lease market in short: in China all gold leases are settled through the SGE (there can be an off-SGE lease market, but it would be highly illiquid). Both lessor (lender) and lessee (borrower) are required to have an SGE Account. If a lease is agreed between two parties gold is transferred from one SGE Bullion Account to the other, when the lease comes due the gold is returned. At SGE level it’s as simple as that.
There is a big difference between jewelers that lease gold in contrast to miners and speculators. Jewelers lease gold because they need physical gold for fabrication; miners and speculators lease gold because they are seeking cheap funds, they will always sell spot the leased gold (without withdrawing the metal) at the SGE to use the proceeds. Why would a speculator withdrawal the metal?
Therefor, if SGE withdrawals capture leased gold this is for genuine jewelry fabrication that eventually ends up at retail level. When a jeweler needs to repay the lease it simply buys gold at the SGE to subsequently transfer it from its SGE Bullion Account to the lessor’s SGE Bullion Account. It’s not likely a jeweler would buy gold off-SGE to repay a lease, which then would need to be refined into newly cast bars by an SGE approved refiner to enter the SGE vaults. Gold leasing by jewelers can increase SGE withdrawals (for genuine gold business) but not so much supply to the SGE.
… No statistics are available on the outstanding amount of gold tied up in financial operations … but Precious Metals Insights [PMI] believes it is feasible that by the end of 2013 this could have reached a cumulative 1,000t…
This 1,000 tonnes figure is based on a misunderstanding regarding the Chinese gold lease market. PMI assumed there was 1,000 tonnes of gold tied up in financing deals based on the yearly lease volume in China, which was 1,070 tonnes in 2013. However, the yearly lease volume is not the gold that is leased out at any point in time, but reflects the aggregated volumes disclosed on all lease contracts that are executed over one year’s time in the Chinese domestic gold market (turnover). Meaning, if 100 mining companies lease 2 tonnes of gold for 1 month in 2016 and all leases are rolled over 4 additional months, the yearly lease volume would be 1,000 tonnes (100 x 2 x (1 + 4)), while on 31 December 2016 the total amount of gold leased out could be nil. (It’s impossible there was 1,000 tonnes used in round tripping as gross export from China has never been more than 330 tonnes)
In addition, the WGC used the words ‘tied up‘ for the gold used in financing operations, which sounds as if the market will be flooded when the gold is untied. The words ‘tied up’ can be misleading, let me explain: If a speculator borrows gold he will promptly sell it spot, this gold will not leave the SGE system. During such a lease period there is nothing tied up, there is just a debt to be repaid. When the lease comes due the lessee has to buy gold in the market (SGE) to settle the debt, which is the opposite of what the WGC insinuates what happens when gold is untied. In case a jewelry company leases gold the words tied up are more appropriate, in my view, as the borrowed gold bars are in transit from being processed to being sold as jewelry. Gold involved (tied up) in these leases can only be a share of the total amount of gold leased out in any point in time, because we all agree most leases in China are done for financing. There is only a small percentage of total gold loans tied up by jewelry companies.
… a good part of the withdrawals represent gold that is used purely for financing and other end-uses that are not equivalent to real consumption.
Needless to say I don’t agree for the reasons just mentioned regarding gold leasing (speculators and miners borrowing gold will never withdraw the metal). Am I the only one? No. When Na Liu of CNC Asset Management Ltd, visited the SGE in May 2014 he spoke to the President of the SGE Transaction Department. From Na:
First, the withdrawal data reflects the actual gold wholesales in China. In 2013, the total gold withdrawal from the SGE vaults amounted to 2,196.96 tonnes. The President of SGE Transaction Department (The President) said: “This 2,200 tonnes of gold, after leaving our vaults, they entered thousands of Chinese households in the form of jewellery and investment purchases.”
… Second, none of the 2,200 tonnes of gold was bought by the Chinese central bank. The President said: “The PBOC does not buy gold through the SGE.”
… Third, the financing deals do not exaggerate SGE’s assessment of China’s gold demand. This is because “the financing deals do not take place after the gold leaves the vaults.”
The President of the SGE’s Transaction Department is clearly stating most leasing happens within the SGE system and this metal is not withdrawn. Therefor, gold leasing by speculators does not inflate SGE withdrawals and thus does not explain the difference between SGE withdrawals and Chinese consumer gold demand as disclosed by the GFMS.
Remarkably, when I asked the WGC about the details in 2014 they replied [brackets adde by Koos Jansen]:
Gold leasing: Banks have built up this business to support China’s burgeoning gold industry. Miners, refiners and fabricators all have a requirement to borrow gold from time to time.For example, fabricators borrow gold to transform into jewelry, sell and then repay the bank with the proceeds. It is an effective way for the fabricator to use the bank’s balance sheet to fund its business. Banks have strict policies in place for who they can lend to, and these have been tightened over recent years, but during PMIs field research it identified that, in some instances, organizations other than genuine gold business had used this method to obtain gold, which it would then sell to obtain funding [in this case the gold wouldn’t be withdrawn from the SGE vaults]. It would then hedge its position. According to PMI, this can generate a lower cost of funding than borrowing directly from the bank. Our colleagues in China think this would be a very small part of total gold leasing; the majority of it would be used to meet the demands of genuine gold businesses.
In their email the World Gold Council admits gold leases that are withdrawn from the SGE vaults are used for genuine gold business and being part of true gold demand. This is more confirmation gold leasing cannot explain the difference.
Over recent years we have observed a rising number of commercial banks participating in the gold leasing market. … It’s estimated that around 10% of the leased gold leaves the SGE’s vaults. The majority is for financing purposes and is sold at the SGE [and stays within the SGE vaults] for cash settlement.
In conclusion, (i) round tripping gold flows are completely separated from the Chinese domestic gold market (SGE) and therefor cannot have caused the difference. In addition, (ii) gold leasing only inflates SGE withdrawals when used for genuine gold business and therefor cannot have caused the difference either.
Last week the Shanghai Gold Exchange (SGE) launched a new English website to offer international customers more information and tools on trading gold in renminbi through its subsidiary in the Shanghai Free Trade Zone the Shanghai International Gold Exchange (SGEI). BullionStar took the opportunity to translate a speech by a Teng Wei, Deputy General Manager of the SGEI, named “How China’s Gold Market Can Help The RMB Achieve International Status” that was held at the Renminbi World summit in Beijing on the 29th and 30th of November 2016. In the speech Teng Wei outlined his vision for the SGEI going forward regarding renmibi (RMB) internationalization, connecting the onshore and offshore renminbi market and increasing gold market share.
My comment before you read the translation:
1) In the financial blogosphere the general perception is that the SGEI has been a failure since it was launched in September 2014. This analysis is based on the assumption that the trading volume of the most popular SGEI contract (1 Kg 9999 – iAu99.99) has been tepid for two years now. But this analysis neglects two important elements.
First, iA99.99 can be traded competitively “on Exchange”, but also in the OTC market. The OTC possibility is hardly known by commentators in the English world, though the related volumes are significant. Have a look at the next chart in which I’ve plotted iAu99.99’s weekly trading volume “on Exchange” and in the OTC market. Clearly iAu99.999 is traded mainly in the OTC market.
Second, international customers of the SGEI can not only trade the SGEI gold contracts, but they can also trade SGE (domestic) gold contracts. Logically, as at present liquidity on the SGE is much higher than on the SGEI, many international customer that seek to trade gold in renminbi, and don’t need to export the metal, will choose to trade SGE gold contracts.
When observing total trading of all SGE(I) gold contracts, there is a clear rise in volume since the SGEI was launched.
Up till now international customers are mainly trading SGE contracts. The significant rise in trading volume of all SGE(I) contracts since September 2014 is due to the inception of the International Board (SGEI). In the second week of November 806 tonnes was traded on the SGE(I), the highest amount ever.
So the launch of the SGEI has not been a failure in my opinion – it has elevated gold trading in (offshore) renminbi.
2) Teng Wei mentions that in 2015 gold demand in China and India was 985 and 849 tonnes respectively. In the case of China this refers only to consumer demand, not institutional demand. Chinese consumer and institutional demand in 2015 combined was well north of 2,000 tonnes.
3) A gold exchange doesn’t flourish overnight. The SGE was launched in 2002; in that year its total trading volume was 22 tonnes and withdrawals accounted for 16 tonnes. Ten years later total trading volume was 3,175 tonnes and withdrawals accounted for 1,138 tonnes. In 2015 total trading volume was 17,033 tonnes and withdrawals accounted for 2,582 tonnes. The development of the SGE, becoming the largest physical gold exchange globally, took time and it can be no different for the SGEI.
Document Translation [brackets added]:
Teng Wei: China’s Gold Market Opens Up To Boost RMB Internationalization
The 2016 RMB summit was held in Beijing on the 29th and 30th of November. Deputy General Manager of the Shanghai International Gold Exchange Center Teng Wei participated in the forum and discussion on “How China’s Gold Market Can Help the RMB Achieve International Status”. He expressed that using Shanghai’s free trade zone status, investors can open trading accounts denominated in RMB and participate in trading directly through the Exchange’s international board [SGEI] that allows access to most of the precious metal products that are traded in China. The international board has developed relatively well since establishment with active participation from international members and steadily increasing trading volume.
Gold on the international board is quoted and settled in RMB, which effectively connects the RMB onshore market and offshore market. This will extend the scope of RMB usage across borders and provide a new channel for inward capital flows. It is a move that is beneficial to expand the RMB usage to steadily promote internationalization of the RMB.
The actual speech:
Ladies and gentlemen, good afternoon, I am Teng Wei from the Shanghai International Gold Exchange. I am delighted to participate in this forum organized by the Asian Bankers Association to have a chance to speak and interact with everyone about opening China’s gold market to the world and how that can help the internationalization of the RMB.
This afternoon, I would like to touch upon on three topics. The first topic is the new pattern of the internationalization of the RMB and the global gold market. China’s gold market was established in 2002 with the launch of the Shanghai Gold Exchange. If anyone is familiar with the history of China’s gold market, you will know that before the year 2002 the Chinese gold market was entirely ran by the People’s Bank of China, including the process of purchasing, allocating and storing of gold. There wasn’t a single unified market where all participants could trade at the same time. Since the year 2002, with approval of the State Council, the People’s Bank Of China developed gold spot trading on the Shanghai Gold Exchange, as well as gold futures trading and over-the-counter trading via commercial banks, etc, which formed the basis for a multi-level diversified gold market system. While the Chinese gold market was developing rapidly, the pattern of the global gold market was also having a dramatic change. As time passed, the international gold spot market was heavily concentrated in London and the international gold futures market has been concentrated in New York. However, in recent years, with the rise of gold demand in China, India and other Eastern nations, and with the exit of European and American banks from the precious metals market, it’s clear that Western gold is moving to the East. In 2015 gold demand in China and India was at 985 tons and 849 tons respectively. These figures alone account for 45% of global [consumer] gold demand. With gold demand from other markets dipping to various levels, China is not only the world’s largest gold producer and importer of gold, but has also become the world’s largest gold consumer.
Just now, I mentioned that the two main centers for gold trading are London and New York, and the current situation is Western gold flowing to the East. Everyone, have a look at some statistics that I have here, showing that just China and India alone make up over 45% of global gold demand. This was last year’s data.
Since the year 2005, when the RMB exchange rate was reformed, international investors’ willingness to trade in RMB denominated assets has also increased. This has objectively enhanced the Chinese gold market’s international status and garnered attention. In recent years, the RMB exchange rate is expected to have some changes.
The Shanghai Gold Exchange provides the important infrastructure for China’s gold market. ECB officials have mentioned that an important part of promoting the internationalization of the RMB is having a good financial market infrastructure. The exchange is also an important “all-in-one” foundation for gold transactions, clearing, delivery and storage. It serves with the commitment to provide gold investors with efficient and convenient market services. It has been 14 years since establishment of the exchange in the year 2002 and development has been rapid with annual trading volumes increasing 40% on average.
At the end of 2015 there were over 8.6 million individual accounts, over 10,000 institutional accounts and the total gold trading volume for the year reached 17,000 tonnes. The exchange was ranked as one of the largest and we firmly grasped an important opportunity for the internationalization of the RMB with the profound changes happening in the gold market. At the same time, we want to build a harmonious ecological gold market that sets a new path for the global gold market and achieve the status of being a global gold power from a large gold holding nation.
For the second point, I would like to explain how opening up China’s gold market externally to the world can help the internationalization of the RMB. To further promote and innovate China’s gold market, on 18th September 2014, the Shanghai Gold Exchange set up an international board [SGEI], open directly to foreign investors. This move has effectively connected China’s domestic gold market and the international gold market. Using Shanghai’s free trade zone, investors can open trading accounts denominated in RMB and participate in trading directly through the exchange’s international board that allows access to most of the precious metal products that are traded in China. The international board has developed relatively well since establishment with active participation from international members and steadily increasing trading volume.
As of now, the exchange has 67 international members, including most of the world-renowned gold suppliers and traders like Mr Thomas McMahon, who is also our Exchange’s member. At the end of the third quarter of 2016, the international board had recorded a total of 7,837 tonnes of gold traded, with a turnover valued at nearly 200 billion RMB. The Shanghai International Gold Exchange is the test pilot and pioneer for opening up China’s gold market to the world. It is significantly important for further increasing the capacity, expansion and international influence of China’s gold market. In addition, the international board uses RMB for settlements, producing an effective convergence of the RMB offshore and onshore markets, expanding the cross-border use of the RMB and providing a new channel for return of funds. All these points steadily promote the internationalization of the RMB and serve as a useful exploration.
For RMB denominated gold products to gain popularity outside of China, we think the prerequisite is to provide a fair offering for global gold market transactions, with reliable gold benchmark pricing in RMB, using the Shanghai Gold Exchange benchmark pricing mechanism [Shanghai Fix] for our trading platforms. The weight of the gold traded is 1 kilogram, with a fineness of no less than 99.99%. Using a price inquiry method and market volume, a balance is reached to form the benchmark price of gold measured in RMB. The price announcements will be released externally each trading day at 10:15 and 14:15.
At present, the Shanghai gold benchmark price is being used by domestic gold producers and suppliers for hedging and settlements. More and more commercial banks are also using the Shanghai gold benchmark price for gold leasing and financing as the standard. More and more products linked to the Shanghai gold benchmark will be made available.
Other than domestic usage, the Shanghai gold benchmark price is also being actively studied more and more by external markets regarding its application. In October, the exchange signed an agreement with Dubai for the right to use the Shanghai gold benchmark price and authorization was given for the Dubai gold exchange to use the Shanghai gold benchmark price as the standard for offshore RMB denominated futures. The signing of this agreement marks the use of the Shanghai gold benchmark price in international financial markets for the first time. This greatly helps to elevate the international influence of the exchange in global markets and improves the image and reputation of the RMB abroad.
For the third point, I would like to share with everyone how the Shanghai Gold Exchange acts as an important infrastructure for internationalization in three steps. As the forerunner for opening domestic markets and innovation, the Shanghai Gold Exchange cannot forget its historical mission. We are determined to take the international and market-oriented strategy.
Overall, for the internationalization process, we have three steps to take. The first step is to be open and inclusive, actively inviting foreign investors to come in. Just now, we have introduced our international board after the establishment of the Exchange and we will continue to increase publicity efforts. In accordance to high standards and multifaceted principles, we will continue to increase and expand international membership of the Exchange. Accordingly, we have carried out a variety of promotional activities in major financial hubs and countries and regions along the new Silk Road to allow more international market participants to hear the sound coming from the Chinese gold market. The exchange also takes the opportunity to actively learn from the experience of advanced international markets in the optimization of various trading systems and innovation of all kinds of trading products.
For the second step, since we have invited guests inwards, we also have to step outwards. Through cooperation and win-win situations, the gold Exchange can be promoted and step out of China. The Shanghai gold benchmark price has now taken a first step with the Dubai Gold Exchange agreement. This can be considered an ice-breaking move and serve as a cooperation model for other overseas markets and improve the recognition, branding and acceptance of the Shanghai gold benchmark price. Taking this as an opportunity, the Shanghai Exchange, together with the Chicago Mercantile Exchange (CME, COMEX), the Hong Kong Stock Exchange, the Malaysian Stock Exchange and a number of foreign exchange markets explored on long-term cooperation mechanisms that will allow foreign investors to directly participate in our gold market, in RMB denominated gold trading, standard gold settlement, and many other aspects and modes of cooperation that will increase the Shanghai Gold Exchange’s international market influence.
The third step is to realize RMB internationalization and increase global transaction on the exchange through integration and upgrades. As the international financial markets continue to merge and develop, market boundaries are increasingly blurred and we believe that market fragmentation will be removed gradually. In recent years, we can all notice that there are more and more mergers and acquisitions among major exchanges in the world. We hope to learn from the experiences of such joint stock mergers and acquisitions between global exchanges and explore the different modes of industry integration with overseas exchanges. By offering a wide range of local and overseas products through an open platform [SGEI], we hope to create a world class exchange group. The journey of the internationalization of the Shanghai Gold Exchange will epitomize the opening of China’s financial markets to the outside world and play an important part in the internationalization of the RMB. With Shanghai becoming the third most important market in the world after London and New York, the Chinese gold market will make a great contribution to the internationalization of the RMB. Thank you everyone.
From the moment Donald J. Trump got elected as the next President of the United States, on November 8, 2016, the price of gold tumbled 8 % in the remainder of the month – from $1,282 USD/oz to $1,178 USD/oz. Usually these cascades in the gold price go hand in hand with physical sell-offs in the West and strong demand Asia. It appears November has been no exception. The volume of physical gold withdrawn from vaults of the Shanghai Gold Exchange (SGE) in November accounted for 215 tonnes, the highest amount in ten months. Year to date SGE withdrawals have reached 1,774 tonnes.
There have been rumours in the gold space about the People’s Bank Of China (the PBOC) curbing gold import into the Chinese domestic market in response to capital flight. Although my sources have confirmed these rumours, Chinese gold import in November was still very strong at an estimated 140 tonnes. I don’t expect the PBOC will halt gold import all together.
The first mention of the rumour was by Reuters on November 25. By then the premium on physical gold trading at the SGE, which more or less reflects the strength of local demand versus international supply, had reached 2 % (from ~ 0.2 % on November 1). Reuters wrote:
“While we don’t have the exact numbers, we hear that they (Chinese government) have limited the number of importers,” said Dick Poon, general manager at Heraeus Precious Metals in Hong Kong.
In a previous blog post I stated the quote from Poon was not likely to be accurate, because there are 15 banks that have PBOC approval to import gold, but for every shipment a new License must be requested at the central bank. This protocol is referred to as “one batch one License”. Bullion cannot cross the Chinese border without a License. From the PBOC:
There shall be one Import … License of the People’s Bank of China for Gold … for each batch … and the License shall be used within 40 work days since the issuing date.
If the PBOC desires to curb gold import it can simply hand out less Licenses to approved banks, instead of deleting banks from the approved list. The former has happened as far as I can see. The next mention was by the Financial Times on November 30 [brackets added]:
Some banks with licences [approval] have recently had difficulty obtaining approval [Licenses] to import gold, they said — a move tied to China’s attempts to stop a weakening renminbi by tightening outflows of dollars, the banks added.
Although the Financial Times exchanged the terms “approval” and “License”, this is what I thought that was happening: banks are obtaining less import Licenses from the PBOC, which is obstructing supply, pushing up the SGE premium.
Either way the PBOC effort has not severely impacted the volume of Chinese gold demand as SGE withdrawals set a ten-month record at 215 tonnes in November, up 40 % from October. Premium or no premium the Chinese still ‘accumulate on the dips’. Additionally, mainlanders buy gold in Hong Kong where jewelry is cheaper as it doesn’t enjoy VAT. From Live Trading News we read:
“Gold sellers in Hong Kong, where mainland Chinese often buy gold, report an increase in purchases, …” according to published reports. “Some of the buying is also because of the Lunar New Year period next month, a time when buying normally picks up.”
How much of SGE withdrawals were supplied by import? Let’s make an educated guess. In the first nine months of 2016 SGE withdrawals accounted for 1,407 tonnes and China net imported 908 tonnes over this period, implying 65 % of SGE withdrawals was imported. If we use the past months as a reference China imported 140 tonnes of gold in November (=0.65*215). Year to date (-November) China has imported and estimated 1,147 tonnes.
An other possibility would be that elevated SGE withdrawals in November were supplied by scrap and disinvestment from within China (domestic mine output is fairly constant at 38 tonnes per month). Though this is not very plausible because the renminbi gold price went down in November (red line in exhibit 1). Normally scrap supply increases on a rising gold price. And hence, I assume the majority of SGE withdrawals in November were supplied by imports.
There have been concerns in the gold community with respect to a full stop on Chinese gold import. In my humble opinion the PBOC will not completely block imports for a number of reasons:
Despite the rumours of obstructed imports SGE withdrawals were strong in November.
The PBOC hasn’t released an official statement to curb imports.
The PBOC has just spent decades to develop the Chinese gold market in order to strengthen the Chinese economy and internationalize the renminbi. Why cancel the project for problems that can be solved differently?
Curbing gold imports would improve China’s current account. But China has a current account surplus; the capital account is in deficit. Why doesn’t the Chinese government tighten the capital account? In Q3 2016 China’s capital flow was minus 71 billion US dollars. In the same quarter gold import was valued at an estimated 13 billion US dollars. The problem is in the capital account.
SGE premiums started to rise on November 8 exactly when the gold price went down (which SGE premiums often do when the price goes down, exhibit 5). So are these elevated premiums of late fully caused by curbed imports, or simply strong demand? It’s probably a mix of both; in any case there is no full stop on imports. What probably happened is that imports exploded when the price tanked after November 8. As a result the PBOC decided to block shipments.
Most gold analysts surmise COMEX 100-ounce gold futures contracts (GC) can only be physically settled through taking and making delivery. This is technically true when excluding the possibility of EFP trading in GC through the over-the-counter (OTC) market. While on Exchange trading in GC is “executed openly and competitively”, trading GC in the OTC realm (and thus the price of the gold, its form and location) is a “privately negotiated transaction” between buyer and seller. The COMEX is a subsidiary of CME Group, which offers its clients OTC trading on a platform called ClearPort.
Because the COMEX in New York is the most liquid gold futures exchange globally – offering precious metals futures denominated in the world most used currency the US dollar, gold industry participants use GC for a variety of reasons, including hedging metal held outside the contract’s deliverable geography. Subsequently, the contracts can be physically “settled” anywhere at any price through EFP.
In EFP two parties sign a futures contract (short and long) and simultaneously execute a reverse spot transaction (buy and sell). One side sells short the futures contract and buys spot gold (the spot leg is referred to as the related position by CME), while the other buys long the futures contract and sells the related position. EFP trading can increase the open interest, decrease the open interest, or not change it, depending on the existing positions held by both parties before they enter into an EFP transaction. When EFP decreases the open interest the phrase “settle positions” is applicable. Another way of saying it would be “offsetting positions” or “netting out positions”.
This example matches my previous one regarding EFP (hedging metal held outside the contract’s deliverable geography). Any bullion bank, miner or refinery can sell short on COMEX and when the gold needs to be physically “settled”, for example in Switzerland, the short position can be unwound through EFP. The only requirement is that “the quantity of the related position component … must be approximately equivalent to the quantity of the Exchange component” – meaning the spot leg must be more or less 100-ounces of gold, which is the underlying asset of GC. In this example the GC short holder connects through CME ClearPort to Exchange For Physical. In the EFP transaction he will buy long a futures contract and simultaneously sell spot. His long and short will then be netted out while he sells spot the physical in Switzerland. Effectively, a COMEX short has been physically settled outside the contract’s deliverable geography. Naturally, a long position can also be unwound in Switzerland, which is then the other side of the trade.
EFP Moves Kilobars Through CME’s Hong Kong Vaults
In March 2015 CME launched a Gold Kilo Futures contract (GCK) physically deliverable in Hong Kong, but ever since implementation there has been poor participation in this instrument. From the start GCK trading volume has been close to nothing and deliveries rarely occur. Notwithstanding, there are massive volumes of kilobar gold flowing through the CME approved warehouse in Hong Kong owned by Brink’s, Inc.. On average 3.9 tonnes per day are withdrawn from this vault, but sometimes daily withdrawals are as high as 20 tonnes.
Because volume and delivery for GCK on Exchange is so low, the withdraws must be explained by OTC trades. A CME representative actaully confirmed this to me; the physical movement through the Hong Kong vaults is caused by EFP transactions.
But if we look at the GCK volume page we can never observe any EFP trades being disclosed. In contrast, EFP volume of GC is substantial. Can it be gold kilobars in CME’s approved warehouses in Hong Kong are used to settle the 100-ounce futures contracts? Yes.
My theory is that kilobars bought by bullion banks in the West, for example at Swiss refineries, to be consigned to China are hedged on the COMEX and once the gold arrives is Hong Kong the shorts are unwound through EFP. From there the gold is transported by armored truck to Shanghai Gold Exchange designated warehouses in Shenzhen by Brink’s that has a cross-border logistics license from the Chinese government. Supportive to my theory, see exhibit 3 below. Notice the strong correlation between “gold import into Hong Kong versus kilobars received in CME’s vaults” and “re-export from Hong Kong versus kilobars withdrawn from CME’s vaults”.
The correlation points out most gold moving through Hong Kong, which is headed for China, is in kilobar form and moves through CME’s vaults. And because most of this throughput is EFP related, I assume the kilobars are used to settle COMEX futures.
It’s hard to test if my theory is accurate because EFP transactions are executed in the OTC realm and little information is available. Possibly, al throughput in Hong Kong is EFP related but doesn’t impact the GS open interest. If anyone has a different theory please comment below.
London Gold Offsets COMEX Futures
We’ve established gold in Switzerland and Hong Kong is used to “settle” gold futures. But there is also proof gold in London is used to phase out positions on the COMEX. When researching this topic I reached out to William Purpura who is, inter alia, Chairman at Northport Commodities, member of the COMEX Governors Committee, and previously traded on the COMEX floor from 1982 to 2007. I asked Purpura for an example of how EFPs are used. He replied [brackets added by me]:
Most of it [EFP] is done by bullion banks. … It’s mainly for netting out. Lot’s of times London versus New York. You see lots of EFPs posted around 8am in New York on COMEX.
There it is, “London versus New York”, and, “netting out”. From this quote we learn loco London gold is used to execute EFPs to wash out New York futures positions. One can argue the related position in London is “unallocated” – I’m not sure. In the latest formulation by CME on EFP (Market Regulation Advisory Notice RA1311-5R) it’s stated:
Where the related position component … is a physical transaction … the transaction should be submitted for clearing as an EFP transaction type.
Often in wholesale gold market parlance physical is also used for “unallocated gold”, which is not exactly physical in my opinion.
I’m sure there are many more methods than I’ve mentioned to use EFP, or any other privately negotiated transaction (PNT) available on ClearPort, that influences the open interest at the COMEX. One thing is for sure, conventional delivery is not the only way to terminate futures positions. In the gold futures rulebook this is explicitly noted by CME Group. The excerpt below is about terminating a gold futures contracts [brackets added by me].
113102.E. Termination of Trading
No trades in Gold futures deliverable in the current month shall be made after the third last business day of that month. Any contracts remaining open after the last trade date must be either:
(A) Settled by delivery which shall take place on any business day beginning on the first business day of the delivery month or any subsequent business day of the delivery month, but no later than the last business day of the delivery month.
(B) Liquidated by means of a bona fide Exchange for Related Position [/EFP] … .
This is important for our comprhension of the global paper and physical gold market. COMEX gold futures delivery statistics are not all there is to it.
What came to light as on odd discrepancy between GFMS’ Chinese gold demand and “apparent supply” has proven to be a tenacious cover-up by the oldest consultancy firm in the gold market. And not only does GFMS publish incomplete and misleading data on Chinese gold demand, all its supply and demand data is incomplete and misleading. As a result, the vast majority of investors across the globe has been brainwashed to believe total gold supply and demand mainly consists of global mine output and jewelry demand. In reality, the supply and demand data GFMS publishes is just the tip of the iceberg. But the firm is reluctant to admit this publicly, lest their business model would be severely damaged.
GFMS has denied all allegations about their incomplete Chinese gold demand statistics by continuously making up false arguments. Therefore, BullionStar will debunk, once more, such arguments spread by GFMS – which are supposed to explain how from January 2007 until September 2016 the difference between GFMS’ Chinese gold demand and apparent supply reached over 4,500 tonnes – in order to expose true Chinese gold demand.
Since 2013 I’ve witnessed GFMS shamelessly present nine arguments in their Gold Survey reports, but along the way abandoned the arguments that I had debunked on these pages. Indeed, few of all these arguments have ever proven to be valid, illustrated by the fact that GFMS perpetually keeps making up new ones. What’s left is to disparage are the final three arguments from GFMS’ most recent annual report: the Gold Survey 2016 (GS2016). Because GFMS chooses their arguments to be ever more complicated, I’ll have to be precise in my wordings not to allow any margin for interpretation errors. For detailed information regarding the mechanics of the Chinese gold market and supply & demand metrics readers can click the links provided.
Debunking Final GFMS Arguments
In the Chinese domestic gold market nearly all supply (import, mine output, scrap supply) is sold through the Shanghai Gold Exchange (SGE), and so Chinese wholesale gold demand can be measured by the amount of gold withdrawn from the SGE vaults; data published on a monthly basis. As I’ve been reporting on withdrawals from the Chinese core exchange since 2013, the debate between me and Western consultancy firms like GFMS with respect to true Chinese gold demand has centered around these infamous SGE withdrawals (exhibit 1). Per mentioned above, GFMS has put out nine arguments in recent years explaining their reader base why SGE withdrawals do not reflect gold demand. Firstly, let us have a look at the five arguments now abandoned by GFMS:
Chinese commercial bank assets to back investment products. “The higher levels of imports, and withdrawals, are boosted by a number of factors, but notably by gold’s use as an asset class and the requirement for commercial banks to hold physical gold to support investment products.” (Gold Survey 2015, page 78).
Defaulting gold enterprises send inventory directly to refiners and SGE (Gold Survey 2015 Q2, page 7)
No need to discuss these anymore, as GFMS dedicated a full chapter in the GS2016 report titled, “A Review And Explanation Of How China’s SGE’s Withdraw Numbers Are Impacted By Other Trading Activities”, in which the arguments above are not listed, implying GFMS ceased to recognize them as relevant. However, there are three new arguments listed, and one old one, that will be discussed in this post:
Gold leasing activities and arbitrage opportunities (in China gold is money at lower cost) (Gold Survey 2016, page 57, Gold Survey 2015, page 78)
Because gold leasing is an old argument it will only briefly be addressed here.
1. Tax Avoidance
This argument entails an illegal Value-added tax (VAT) invoice scheme. Although this scheme exists, it can not have the impact on SGE withdrawals like GFMS wants you to think.
GFMS introduces its special investigation chapter by stating:
The first and foremost factor behind why we believe the SGE’s withdrawal number differs from the country’s total gold demand is related to China’s current tax system, with some people exploiting this grey area.
… the number of industry participants mushroomed in 2014 and 2015 as other traders became aware of the potential loophole.
The GFMS team uses the terms “tax avoidance” and “loophole”. For the ones that don’t know, tax avoidance and tax evasion are two opposing practices. Tax avoidance is the legal usage of a tax regime to one’s advantage in order to reduce the amount of tax payable by means that are within the law (Wikipedia). Tax evasion is the illegal evasion of tax payable (Wikipedia). In other words, tax avoidance is legal while tax evasion is illegal. In the introduction the GFMS team pretends the tax scheme is legal, while this is anything but true. In China one can risk life imprisonment or the death penalty when caught for tax evasion:
Whoever forges or sells forged special invoices for value-added tax shall, if the number involved is especially huge, and the circumstances are especially serious so that economic order is seriously disrupted, be sentenced to life imprisonment or death and also to confiscation of property.
Then, to add to the confusion, further down the GFMS team writes, “of course, all of the activities are considered illegal by the Chinese government.” Maybe GFMS doesn’t understand the difference between tax avoidance and tax evasion, two diametrically different practices, which makes their professionalism highly questionable.
GFMS writes, “the first and foremost factor behind why we believe the SGE’s withdrawal number differs from the country’s total gold demand is related to China’s current tax system”. So we’re supposed to believe that after all these years – GFMS is operational for decades – and all that has been written on the Chinese gold market, now GFMS has finally found the “first and foremost reason” why SGE withdrawals do not reflect demand? Or did it recently stumble upon this scheme to use in its defence? I think the latter.
Regarding using VAT invoices for tax evasion, the GFMS team must have read this news article by the Shenzhen Municipal Office. In the news, a company called Longhaitong used SGE VAT invoices for tax evasion. How does it work? For example: the prevailing spot gold price on the SGE is 234 CNY/gramme. Company X tells a mom-and-pop jewelry fabricator that they can supply good quality cheap gold, say the SGE spot price minus 2 CNY/gramme, but without a VAT invoice. The mom-and-pop fabricator wants to buy 1 Kg so it gives 232,000 CNY to company X (the mom-and-pop shop will fabricate jewelry from the gold to be sold covertly without VAT to consumers). Company X buys 1 Kg of gold on the SGE at the spot price of 234 CNY/gramme, paying 234,000 CNY. Then company X gives the gold to the mom-and-pop fabricator but keeps the VAT invoice. Up till now, company X has incurred a loss of 2,000 CNY (bear in mind, because of China’s VAT system buyers pay the spot price at SGE which doesn’t include any VAT, but when companies withdraw the metal they receive a VAT invoice from the tax authority that describes 17 % of the all-in price is VAT, because the gold leaves a VAT exempt environment). However, company X can then sell the VAT invoice for 4,000 CNY to, in example, a brick trader. Company X effectively makes 2,000 CNY. If the brick trader alters the subject header on the invoice from “gold” into “bricks” he can tax deduct 34,000 CNY (234,000 / (1+17%) * 17%) from his VAT payable. In this scenario, the brick trader effectively makes 30,000 CNY (34,000 CNY minus the 4,000 it paid to company X). Naturally, all exemplar numbers can vary.
For sure this illegal VAT scheme exists and has been used. But, only to a limited extent – in my conclusion I will tell you the upper bound. Mind you, in the scenario I just described the gold does meet demand, albeit through an illegal scheme!
In addition, the discrepancy between the GFMS Chinese demand figures and SGE withdrawal numbers first appeared in 2008, and have exploded since 2013.
In the GS2016 GFMS writes:
We initially became aware of the scheme in 2013 when it first emerged, but based on information gathered from our contacts, the number of industry participants mushroomed in 2014 and 2015 as other traders became aware of the potential loophole.
The GFMS team wants readers to believe that it was the tax scheme that caused the discrepancy between GFMS Chinese demand and SGE withdrawals since 2013, but the VAT regulation regarding gold has remained unchanged since 2002. Is it believable that criminals found the possibility of these illegal practices 11 years later, exactly when Chinese demand exploded? No. If you click this link, you will see a similar incident that happened in 2010 and was reported at the end of 2011. The VAT scheme has existed for many years and crime incidents happen, but not like GFMS wants you to think.
If the GFMS team was indeed aware of the illegal practices as late as 2013 and thought that was the year when these practices first emerged, then GFMS is not properly informed in the Chinese gold market.
More from GS2016:
One of our contacts with some understanding of this activity estimated that just from Shenzhen alone, such trading activities could have possibly impacted the SGE’s withdrawal volumes by a few tonnes per day. Approximately half of the gold being sold in the black market at discounts would eventually flow back to the SGE.
In my opinion this is speculation. According to the news available, buyers in the black market are those who want the gold to fabricate jewelry that eventually is being met by true demand. In contrast, GFMS wants readers to believe half of the gold involved in the scheme flows back to the SGE. But bars withdrawn from the SGE vaults are not allowed to re-enter, only if they’re recast into new bars by SGE approved refineries (the gatekeepers of the Chinese chain of integrity). For gold involved in VAT invoice schemes to flow back to the SGE, technically SGE approved refineries would be complicit. Though the SGE conducts a campaign to crack down on such illegal tax activities.
As stated above, the VAT scheme is real, though it can not involve as much gold as GFMS wants you to believe. Unfortunately we can’t compute the exact amount recycled through the SGE through this practice, we can only identify the upper bound, which we’ll do in the conclusion.
As background information: when gold is withdrawn from SGE vaults and promptly flows back to the SGE, this overstates withdrawal numbers as it creates equal demand and supply that has no net effect on the price. Therefore, such recycle flows should not be counted in supply and demand statistics. Readers can click this post for more information.
Financial Statement Window Dressing
The GFMS team writes:
Some companies attempted to build up their revenues by merely trading and withdrawing physical gold from the SGE vault so it would appear they have a high level of business activity, while in reality there is no real genuine demand behind this.
Trading can build up revenues but why do these companies withdraw gold? That doesn’t make economic sense. If a company buys gold on the SGE and leaves the gold in the SGE vault, the gold will be recorded as “inventory” on the company’s balance sheet. If the company then withdraws the gold, the gold is still regarded as “inventory”, so what’s point of withdrawing gold? Changing the location of the gold doesn’t change the accounting nature of the gold.
It is technically possible to buy gold on the SGE, withdraw, refine it into new bars, redeposit the bars into SGE vaults and sell the bars. However, this will incur expenses. When the point is “window dressing”, why incur unnecessary expenses? More logic would be to leave the gold in the SGE system. This argument is false.
Retailers Selling Unsold Inventories Directly to Refiners
In this section, the GFMS team writes:
Retailers often prefer to sell a portion of their working stock at a discount directly to refiners in order to maintain inventories at a desirable level.
Why waste the fabrication costs of jewelry when retailers can sell the products at a discount to customers?
By selling to refiners, even if such a transaction may result in a financial loss, it still counts as revenue; but doing the latter only increases the expense category and provides no benefits to the company’s revenues or asset value.
Let’s assume an unsold jewelry stock is worth of 1,000 CNY. The retailer sells it to a refiner at 800 CNY, which results in a loss of 200 CNY. The inventory item on the retailer’s balance sheet is reduced by 1,000 CNY and the cash item increases by 800 CNY. The net result is that the total asset value of the retailer decreases by 200 CNY, then how can this practice provide benefits to the asset value?
As an example, during a field research trip earlier this year, a local refiner indicated that one jewellery retailer has sold approximately 40 tonnes of unsold jewellery pieces to them in a single two month period.
But this quote doesn’t mention what the unsold jewelry pieces become in the end. Possibly, these pieces become gold wires, which might be used by jewelry fabricators instead of becoming gold bars that flow back to the SGE. GFMS pretend the majority of gold in China is continuously recycled through the SGE, which is not true. Many refineries are note even approved by the SGE to supply gold bars.
Gold Leasing Activities And Arbitrage Opportunities
This argument is one of the oldest and most persistent. But we can be short about this; in the Chinese gold lease market nearly all trades are conducted within the SGE system. Any speculator borrowing gold for cheap funding will not withdraw his metal loan, as his incentive is to sell spot for the proceeds. GFMS fools readers by mentioning high leasing activity, but it neglects to mention leases aren’t withdrawn from the vaults. Only a jewelry fabricator would withdraw borrowed gold because he wants to fabricate products to meet demand. For more information you can read this post on the Chinese gold lease market.
Over recent years we have observed a rising number of commercial banks participating in the gold leasing market. … It’s estimated that around 10% of the leased gold leaves the SGE’s vaults. The majority is for financing purposes and is sold at the SGE [and stays within the SGE vaults] for cash settlement.
This argument is false.
Furthermore, it’s noteworthy that GFMS writes:
From the perspective of the bank, lending physical gold is an off-balance sheet item,…
But as I’ve demonstrated in this and this post the majority of the “precious metals” on the Chinese bank balance sheets reflects back-to-back leasing. Meaning banks borrow gold in the SGE system to subsequently lend out at a higher lease rate. So neither do the Chinese bank balance sheets influence SGE withdrawals. What withdrawals largely reflect are direct purchases by individual and institutional investors at the SGE. True demand.
There is a very limited extent to which the VAT scheme can explain the difference between GFMS’ demand and SGE Withdrawals. I wrote previously that indeed there is certain amount of gold being withdrawn from SGE vaults, which, for various reasons, finds its way back to the SGE in newly cast bars – overstating SGE withdrawals as a proxy for wholesale demand. Unfortunately nobody knows exactly the volume flowing through the SGE that distorts withdraw data. But, we do know the upper and lower bound. The upper bound is the difference between SGE Withdrawals and apparent supply, the lower bound is zero.
GFMS only measures consumer demand (jewelry, retail bar and coin, and industrial demand) and not institutional demand (direct purchases at the SGE). This is not speculation this is a fact, and in China everyone can buy gold directly at the SGE so this explains the immense withdrawals. GFMS is fully aware of this but refuses to acknowledge it – because that would ruin their business model. Instead GFMS pretends that the difference between consumer demand and SGE withdrawals is all caused by gold being recycled through the central Chinese exchange. But how is this possible? If the Chinese gold market would simply be a merry-go-round fest, how come the Chinese import thousands of tonnes of gold that are not allowed to be exported? What GFMS suggests is not possible. The fact China keeps importing reveals demand. Another chart:
Theoretically the upper bound for the VAT scheme to have recycled gold through the SGE equals the difference between SGE withdrawals and apparent supply (the difference in exhibit 4 between the red and center columns). That’s the sole leeway we can debate about. As supply equals demand, demand cannot be lower than apparent supply. I should add, not unimportant, we know GFMS’ scrap supply data does not include disinvestment (institutional selling directly to refineries). So disinvestment must be included in the difference between SGE withdrawals and apparent supply as well. Have another look at exhibit 3. But, because we don’t know the amount of disinvestment, neither do we know the amount of distortion (VAT scheme and other recycling flows).
That’s why in exhibit 1 I’ve disclosed the aggregated difference between apparent supply and GFMS demand. There can be no mistake about this volume, it reflects true demand and it has mushroomed into +4,500 tonnes since 2007. GFSM can present many more arguments in future reports, but it won’t change the fact that true demand is at least equal to apparent supply.
This post is part of the Chinese Gold Market Essentials series. Click here to go to an overview of all Chinese Gold Market Essentials for a comprehensive understanding of the largest physical gold market globally. This post was updated in 2017.
In the Gold Survey 2016 by Thomson Reuters GFMS there is a complex illegal scheme described whereby criminals obtain VAT invoices from the Shanghai Gold Exchange (SGE) for tax evasion. According to GFMS this scheme is one of the reasons why SGE withdrawals are significantly higher than “Chinese consumer gold demand”. To be able to properly clarify this scheme I will expand in this post on the workings of the VAT system in China’s Gold Market. The scheme has certainly existed for years, but not anywhere near the volume and frequency GFMS portrays.
The Current VAT system in China was adopted in 1994 as part of the economic reform and is often regarded as one of the most complex systems in the world. Here the discussion is simplified somewhat, not to get entangled in details that are not important. Be aware this article does not discuss income tax.
The General VAT System In China
China’s VAT is chargeable on the sale of goods, provision of processing and repair services, and the importation of goods. The standard VAT tax rate is 17 %, a couple of household necessities enjoy a preferential 13 % VAT rate. When visiting any shop or supermarket in China, you will never see any VAT disclosed separately from the unit price. In China it’s common practice to show customers VAT-inclusive prices. In addition, all the prices listed on China’s Commodity Exchanges, the Shanghai Futures Exchange, Dalian Commodity Exchange, Zhenzhou Commodity Exchange and Shanghai Gold Exchange, are VAT-inclusive prices. As a result, if you see a notepad computer priced at 3,510 CNY (onshore renminbi) in China, the VAT is 510 CNY and the VAT-exclusive price is 3,000 CNY.
The “VAT-liable entities” are responsible for collecting the VAT and hand in the money to the tax authority. The VAT-liable entities in China are divided into two categories: general VAT taxpayers and small-scale VAT taxpayers. General VAT taxpayers are large firms that have annual sales large enough and the ability to maintain an accounting system sophisticated enough to accurately calculate output VAT and input VAT. Small-scale VAT taxpayers are firms that don’t satisfy these criteria. Since small-scale VAT taxpayers are not relevant to our discussion, so the focus will be put on general VAT taxpayers.
The formula for computing the VAT payable to the tax authority for a general VAT taxpayer is:
VAT payable = output VAT – input VAT
Output VAT = current period taxable sales * applicable VAT rate
Input VAT = current period costs of eligible purchases * applicable VAT rate
Here is a small example to illustrate VAT computing. Suppose company A is a laptop wholesaler that purchases laptops from Dell, and re-sells them to local retailer B. Company A buys 100 laptops from Dell at the VAT-exclusive unit price of 3,000 CNY. The total VAT-exclusive amount for the goods is 300,000 CNY and total VAT is 51,000 CNY.
Dell then issues a VAT invoice on which the 300,000 CNY and 51,000 CNY are recorded.
From Dell’s perspective, the 51,000 CNY is Dell’s output VAT, but from company A’s perspective the 51,000 CNY is its input VAT.
After having bought the laptops from Dell, company A then re-sells them to retailer B at the VAT-exclusive unit price of 4,000 CNY. Implying, the total VAT-exclusive amount for the laptops is 400,000 CNY and the total VAT is 68,000 CNY.
Company A then issues to retailer B a VAT invoice on which the 400,000 CNY and 68,000 CNY are recorded. From company A’s perspective, the 68,000 CNY is its output VAT but from retailer B’s perspective the 68,000 CNY is its input VAT.
At the end of the month, the VAT payable by company A is 17,000 CNY, which is to be paid to the tax authority.
VAT payable by company A = output VAT – input VAT = 68,000 – 51,000 = 17, 000 CNY
Company A has to keep Dell’s VAT invoice safe and demonstrate the invoice to the tax authority. If company A couldn’t produce Dell’s invoice to the tax authority, then the 51,000 CNY wouldn’t be allowed to be deducted and company A would have to pay the tax authority 68,000 CNY.
They are four kinds of receipts and invoices in China we will discuss:
Special VAT invoice (SVI)
Customs office special receipt for the payment of import VAT
General VAT invoice
Shanghai Gold Exchange invoice (SGE invoice)
1. Special VAT invoice (SVI). In order for input VAT to be used as a tax credit to offset the output VAT, the input VAT must be substantiated by a “special VAT invoice” (SVI) or “customs office special receipt for the payment of import VAT”. SVIs are issued when a general VAT taxpayer sells taxable goods and services. General VAT taxpayers must purchase blank SVIs from the tax bureau. In China, entities can’t produce any VAT invoice of their own – or it will be fake – but all transaction need to be recorded through an invoice. An SVI looks like this:
2. Customs office special receipt for the payment of import VAT. A “customs office special receipt for the payment of import VAT” is used for imported goods. Suppose company A buys a computer from Australia at the price 10,000 CNY, assuming no tariffs. The exporter in Australia can never give company A a Chinese SVI, but the imported computer does enjoy VAT. Therefore company A must pay the Chinese Customs Office 1,700 CNY (10,000 CNY * 17 %). Upon receiving the money, the Chinese customs office will issue a “customs office special receipt for the payment of import VAT”. With this special receipt, company A can obtain the tax credit (input VAT) from the imported computer.
3. General VAT invoice. A general VAT invoice looks like this:
A general VAT invoice looks very similar to an SVI, though a general VAT invoice cannot be used in obtaining VAT credits when calculating payable VAT. In other words, if you declare to the tax authority that you have 1,000 CNY input VAT but can only produce a general VAT invoice to substantiate your declaration, the tax authority will not recognize the invoice. General VAT invoices are issued by general VAT taxpayers strictly for accounting purposes when they make sales to consumers that will not use the purchase for input VAT. Effectively, general VAT taxpayers issue SVIs for sales to other general VAT taxpayers, and general VAT invoices for sales to consumers.
4. Shanghai Gold Exchange invoice (SGE invoice). The “Shanghai Gold Exchange invoice”, or SGE invoice, is designed by the Shanghai Gold Exchange under the supervision of the national tax authority. It’s a pity I can’t find an image of a SGE invoice.
VAT Policy For Gold In China
In China, gold is divided in several categories. There are gold,gold products and ore, and gold is subdivided in standard gold and non-standard. Gold is unwrought/unforged gold, like bars and ingots (HS code 7108120000 and 7108200000). Standard gold refers to gold bars or ingots having a fineness of 9999, 9995, 999 or 995, and a weight of 50g, 100g, 1kg, 3kg or 12.5kg. On the Shanghai Gold Exchange and the Shanghai Futures Exchange (SHFE) only standard gold can be traded. Non-standardgold includes any gold that doesn’t satisfy standard gold criteria, in example 200g ingots. Gold products mean semi-finished gold and finished products of gold, like coins, jewelry and ornaments.
When gold producers and gold traders (general VAT taxpayers and small scale VAT payers) sell non-standard gold off-SGE the VAT is exempt. Gold imported into the domestic market (non-standard and standard gold), for the ones that have an import license, is also VAT exempt.
If standard gold is not sold through the Shanghai Gold Exchange (or Shanghai Futures Exchange), a 17 % VAT tax rate will apply. If standard gold is sold through the Shanghai Gold Exchange (or the Shanghai Futures Exchange), then the VAT is exempt. But it’s the invoicing procedure that is quite complex. An example will follow to illustrate the whole process.
Let’s tie everything together. Suppose the following trades occur. ICBC imports 1 kg of Au99.99 (SGE 1 Kg 9999 gold ingot), which is standard gold, from Switzerland at the price of 230 CNY/gramme (around 1,033 USD/oz). ICBC then sells the gold on the SGE at the price of 234 CNY/gramme. Jewelry manufacturer Laofengxiang is on the other side of the trade. Laofengxiang then withdraws the gold, makes it into gold ornaments, and sells all of them to a retailer at the VAT-exclusive price of 300 CNY/gramme. The VAT payable and receipts and invoices of different parties are as follows:
When ICBC imports the gold, it receives a “customs office special receipt for the payment of import VAT”. On this receipt the total VAT-exclusive amount for the gold is 230,000 CNY and the VAT is 0 (ICBC’s input VAT), because imported standard gold is VAT exempt. When ICBC sells the gold at the price of 234 CNY/gramme on the SGE, it needs to issue to the SGE a general VAT invoice, on which it’s recorded 234,000 CNY for the gold and 0 VAT (ICBC’s output VAT), because selling standard gold on the SGE is also exempt from VAT. Upon issuing the general VAT invoice, ICBC will receive an SGE invoice from the exchange.
After ICBC and Laofengxiang have concluded the deal, the SGE will issue both ICBC and Laofengxiang an SGE invoice respectively. After Laofengxiang withdraws the gold from the vault, the tax authority will issue a SVI on behalf of the SGE, which the SGE distributes to Laofengxiang. On the SVI, the total VAT-exclusive amount for gold is 200,000 CNY and the amount of VAT is 34,000 CNY. The tax authority decides these two numbers using the following formula:
The total VAT-exclusive amount for gold on the SVI = SGE transaction price * quantity / (1+17%) * 100 % = 234 * 1,000 / (1+17%) * 100% = 200,000 CNY
The VAT amount for gold on the SVI = SGE transaction price * quantity / (1+17%) * 17 % = 234 * 1,000 /（1+17%）* 17 % = 34,000 CNY
To understand the reasoning behind this calculation, please note that after withdrawing the metal, this standard gold leaves a VAT exempt environment (the SGE system), for an environment that is not exempt from VAT, and hence VAT is born into existence. When Laofengxiang doesn’t withdraw the gold, the SGE invoice is the only invoice it will receive for accounting purposes – Laofengxiang needs some evidence for accounting entries. If it withdraws the gold, then it will receive an SVI because the standard gold withdrawn can be manufactured into new gold products and sold off-SGE. Therefore Laofengxiang will need to claim input VAT. General VAT taxpayers that withdraw will get a SVI, individuals that withdraw from the SGE will not. The input VAT noted on Laofengxiang’s SVI from the SGE is not an amount Laofengxiang paid to ICBC as VAT, but Laofengxiang is allowed to deduct this amount from its output VAT.
As mentioned, after concluding the purchase of 1kg 9999 gold on the SGE at the price of 234 CNY/gramme, Laofengxiang receives an SGE invoice and also an SVI after it withdraws the gold. The input VAT is 34,000 CNY, which is described above. Laofengxiang then fabricates and sells all the gold ornaments made from the 1kg of gold at the VAT-exclusive price of 300 CNY/gramme. Therefore the output VAT is 51,000 CNY.
Therefore, the VAT payable for Laofengxiang is 17,000 CNY. Since Laofengxiang has the SVI from the SGE, the tax authority will accept the 34,000 CNY input VAT as a tax credit to deduct from the output VAT.
Up till now, readers will have a general idea on how the VAT system works in China’s gold market.
Core Supply & Demand Data Chinese Gold Market Q1-Q3 2016
Chinese gold demand is still going strong this year, albeit less than in 2015. The most likely reason for somewhat lower demand has been the strength in the price of gold in the first three quarters of this year, to which the Chinese reacted by subduing purchases. From 1 January until 30 September 2016, the gold price went up 24 % in US dollars per troy ounce, from $1,061.5 to $1,318.1; measured in renminbi the price went up 28 % over the same period.
Now I have proven the gold on Chinese commercial bank balance sheets has little to do with physical gold ownership of these banks, but mainly reflects back-to back leases and swaps, we can be positive that data on withdrawals from the vaults of the Shanghai Gold Exchange (SGE) roughly equals Chinese wholesale demand. For now that is, as future developments can always alter our metrics.
Below is a chart showing withdrawals from the vaults of the SGE and the price of gold in yuan per gram. The most significant trends of recent years are still in effect; in the short term, when the gold price is falling Chinese demand increases (2013 and 2015), when the gold price is rising Chinese demand declines (2016). This trend is supported by SGE premiums that have an inverse correlation with the price of gold, when the price of gold declines, SGE premiums escalate and vice versa – I will show charts below. Furthermore, in the long term we can observe consistent growth in Chinese gold demand due to the opening up and development of the domestic market.
SGE withdrawals in the first three quarters of 2016 accounted for 1,406 tonnes – still impressive – down 29 % from 1,986 tonnes in 2015, which was a record year. Annualized SGE withdrawals are set to hit 1,877 tonnes in 2016.
Notable, “known net import” by China is relatively strong compared to SGE withdrawals in 2016. Total net import in the first three quarters of this year has aggregated to 905 tonnes – annualized 1,206 tonnes – or 64 % of SGE withdrawals, versus an import/withdrawals ratio of 53 % in 2015. As mine supply to the SGE is fairly constant, recycled gold through the SGE must be lower this year than last year. As a rule of thumb, we use the equation:
The largest net exporter to China is still Hong Kong, having transhipped 608 tonnes to the mainland from January until September 2016, up 5 % compared to 2015. The volume Hong Kong exports to the mainland has been quite constant since 2014, while in 2013 China’s special administrative region was a substantial larger supplier.
(There have been rumors that Hong Kong ’s export to China is overstated in the official data by the Hong Kong Census & Statistics Department, caused by fake exports. In the chart below you can see that the share of exports relative to re-exports from Hong Kong to China this year has increased from previous years. Potentially this signals fake exports, as it’s easier to over invoice an export than re-export, though I haven’t found hard evidence for this scheme. When I do I will report accordingly.)
The second largest exporter to China is Switzerland, having supplied a net 229 tonnes so far this year, which is 22 % more than last year. Clearly, direct shipments from Switzerland to China have replaced shipments via Hong Kong.
Direct net exports by the UK to China mainland have collapsed by 92 % this year compared to 2015, from 210 tonnes to a mere 18 tonnes. The reason being, the UK has been the largest net importer globally this year, which is related to the strength in the gold price early this year. UK net gold trade is a proxy for Western institutional supply and demand.
Australia’s direct export to China is down this year as well (in the first eight months, data for September has not yet been released). I’ve computed the data as described in my post Australia Customs Department Confirms BullionStar’s Analysis On Gold Export To China. Following this method, the land of down under has sent 50 tonnes of gold directly to China during the first eight months of this year, down 23 % from 65 tonnes in 2015.
Despite press releases suggesting Russian gold enterprises are strengthening ties with the SGE, I have identified only one shipment of 30 Kg by the Russian Federation directly to China in 2016. In 2013 the Russians directly net exported 50 Kg to China.
Data on gold export from South Africa to China is not publicly available.
Since 2014, when the Shanghai International Gold Exchange (SGEI) was erected, there is a possibility “SGE withdrawals” are inflated by withdrawals from vaults in the Shanghai Free Trade Zone; gold that is allowed to be exported abroad – the free trade zone is not part of the domestic market. But as far as I know any activity on the SGEI lacks foreign enterprises that buy gold to withdraw and export. A couple of months ago a source at a large Chinese bank told me the SGEI is mainly used by Chinese banks to import gold into Chinese domestic market. In addition, I haven’t bumped into any large importers from China. Occasionally India imports a few hundred Kg, but that’s it.
The emblematic difference between “Chinese gold demand as disclosed by GFMS” and SGE withdrawals – displayed in exhibit 7 – is due to GFMS’ incomplete metrics. For decades this consultancy firm has been denying the existence of institutional supply and demand in above ground gold, which is far more important to price formation than retail sales and mine supply, the predominant flows published by GFMS. The essence of this swindle can be read in my blog post The Great Physical Gold Supply & Demand Illusion. I also have a few more blog posts in the pipeline that discuss GFMS’ most recent gold supply and demand data.
I expect November to be a very strong month for SGE withdrawals. Mentioned in the introduction segment of this post, there is a trend in Chinese wholesale gold demand in relation to the gold price. Whenever, the gold price is climbing, Chinese demand is subdued, accompanied by low SGE premiums; when the gold price is decreasing, SGE withdrawals and premiums in China shoot up. The relationship between the gold price and SGE withdrawals can be viewed in exhibit 1. Below in exhibit 8 & 9, readers can see the relationship between “SGE end of day prices and premiums”.
Note, the gold price on the SGE and the premium have an inverse correlation.
I already mentioned that SGE withdrawals in the first nine months of 2016 have been subdued due to a rally in the gold price. However, high premiums at the SGE in November forecast elevated withdrawals for the month. Since Trump got elected on November 9, and price of gold started tumbling, SGE premiums have broken a three-year record. This signals strong demand.
In the next chart from Goldchartsrus.com we can see the premium on the SGE’s most traded physical contract Au99.99 has risen since November 9 and reached 3 % by 24 November. Levels not seen since 2013 (exhibit 8).
Although the relationship between the gold price and SGE premiums has been in place for years, Reuters reports the high premiums in November are caused by worries on import restrictions. From Reuters:
Gold premiums in top consumer China jumped to the highest in nearly three years this week on worries over a supply shortage that traders said were due to Beijing’s efforts to restrict import licenses.
“While we don’t have the exact numbers, we hear that they (Chinese government) have limited the number of importers,” said Dick Poon, general manager at Heraeus Precious Metals in Hong Kong.
To me this statement doesn’t make sense. At this moment that are 15 banks approved by the PBOC to import gold. Limiting the number of importers would cause less importers to import more gold in order to balance the domestic market (supply gold from abroad when necessary). In the Measures for the Import and Export of Gold and Gold Productsdrafted by the PBOC in March 2015 it states:
… An applicant for the import … of gold … shall have corporate status, … it is a financial institution member or a market maker on a gold exchange [SGE] approved by the State Council.
… The main market players with the qualifications for the import … of gold shall assume the liability of balancing the supply and demand of material objects on the domestic gold market. Gold to be imported … shall be registered at a spot gold exchange [SGE] approved by the State Council where the first trade shall be completed.
The Chinese government could lower imports by distributing less “import licences” to approved banks. As, every approved bank still needs to submit for a license for every gold import batch. Logically, lowering imports would be done by the PBOC through handing out less licences.