GLD is the world's biggest repository of gold bars, if you remember the 'country of refiner origin' breakdown, this is the most recent snapshot (data is from a specific document issue - not an aggregate). I have also included the graphic to remind that we're only interested in the 400-oz variety.
And here is Julius Baer Gold Bars, with the countries in the same order as above. Let's be fair, their gold holdings are only a fraction of the size of GLD (but still a significant amount). The 'Global' represents a brand where the name is ambiguous - in this case, a single brand: "METALOR". Originally I thought these might be Metalor Hong Kong, but the serial numbers are not very similar to the Metalor HK bars (more study is needed here).
There are a few reasons I constructed this chart. First of all, you'll notice the higher comparative percentage of Swiss-refined bars which makes perfect sense - their vaults are located in Switzerland and geographically speaking, there will be more Swiss bars available. We also see less North American refiners here, compared with the vaults in London.
Anyway, the fact remains that (despite looking) we haven't found even one Chinese-Refined LBMA gold bar in a major ETF, which still lends support to the idea that China is holding onto its gold refinery output. This is supportive of Freegold theory, and also any other 'Chinese are keeping their Gold' narrative.To put it in perspective, this is a massive hurdle for anyone who still maintains that gold is just a commodity like the others - China has stuff pouring out of factories like a river including silver bars and other metals, so why should LBMA gold bars appear be absent from the list? (especially since we can safely assume they are being produced).
Although the data fits the story, are there any other explanations? In the spirit of discussion, I've tried to list some counter-arguments I can think of - and they are really, really weak - in fact I shall put forward arguments against my own suggestions. Feel free to put more forward, and we can add them to the list. Presented in order of likelihood. I wish I had some solid facts on the whereabouts of their gold, but we can only surmise. A listing of Chinese LBMA-gold accredited refineries can be seen here, and these are the ones we suspect are making lots of those 400oz gold
1. Geography Constrains Movement?: China is a long way away from the London Market and because gold moves around slower (than silver bars for example), not seeing any Chinese bars in America or Europe is perfectly acceptable - all the other bars have had decades to get to their current location. Perhaps the shipping and insurance against losses (pirates) is simply prohibitive and ETF demand can be met entirely with local stocks - as demonstrated in the Swiss-based ETF data above. So generally, we know this to be an existing effect, but we don't yet know the full extent. Against this idea: The number of shipping lanes in the modern world or the ability to get Fedex-style air deliveries within 24 hours if really wanted it - i.e. geography is not the problem that it used to be. Chinese are generally happy to sell in bulk at a razor-thin margin if it outbids other competitors - in theory the market for refinery output should receive equal treatment. ALSO: They seem to have no trouble shipping all the silver bars all the way to London!!! update: "...flying gold is relatively quick and cheap, and a lot cheaper on a % of metal value compares to shipping silver around" (ref: Bron Suchecki, in comments below).
2. Hong Kong is Chinas Gold???: In theory, because Hong Kong has such a small land area, most of it's gold refineries probably get gold from China, which may account for the high percentage of Hong Kong-refined gold, and absence of China. Against this idea: Still doesn't account for all the LBMA-gold accredited Chinese refiners output. Additionally, Hong Kong gold refineries have been operating for a while so most of those ETF bars are going to be old ones anyway (we'll get a better idea once we get a proper date/age/matching system going). update: additionally, Nick Lairds charts show that China is absorbing lots of HK refined gold as well ...
3. Comparatively Late Accreditation???: Most of the LBMA Gold-accredited refineries didn't get accredited until recently so there should be fewer bars compared with other refiners which have been supplying the market for decades. Against this idea: "The Great Wall Gold and Silver Refinery of China" which was LBMA accredited way back in 1981, plus, volume of recent output of Chinese Silver bars suggests those refiners have been really busy, with equivalent output, there should be a lot of Chinese gold bars.
4. Non-Acceptance of Bars???: Perhaps the ETF's prefer not to stock Chinese bars? After all is there not a general perception that Chinese are masters of the knock-off? Against this idea: These refiners are LBMA accredited, plus any fakes would be quickly discovered by the vaults. Silver bars are fine, why not gold? "Retail clients may not like Chinese bars but at the wholesale level no one cares" (ref: Bron Suchecki, in comments below)
5. The bars are being hoovered up by private Chinese investors???: Against this idea: No data at all on this wild conjecture! Not sure what the market would be for 400oz LBMA gold bars, but even if that were large then they would still be competing with the ETF's. Chinese prefer 1kg bars (ref: Bron Suchecki - also advanced by VtC in a previous thread).
------ this next set have been added based on the discussion in the comments below, which I will update in real time as I am able. I can't change the numbering above, so these are presented in the order they were presented. Attributions as indicated.
6. The Chinese are re-melting their bars into smaller 1kg bars??? (raised by
8f6d4ec6-3d96-11e1-bd2e-000bcdcb5194 as per Jim Willie's suggestion) This would certainly explain the absence of 400oz bars, and would be appropriate if China has ambition to alter the trading landscape (ref: Costata). Against this idea: Doing this would automatically exclude their gold from being used for international settlement, unless they plan to display so much clout as to overturn the established standard (which is possible). 'London Trader' is a dubious source, no backing data for the claim, financial institutions trade in 400oz bars as tradition/standard (as observed by Victor) and Chinese tend to follow, not lead. Rationale for having smaller-sized bars is also missing; would be a lot of effort and cost (but it's not like they are lacking the people to do it).
7. Chinese aren't producing London good delivery bars in large quantities??? (suggested by Costata in the Occam's Razor category) i.e. producing them, but not enough to get absorbed into the London Market. Potentially explains most scenarios and allows for normal geographic limitation effect Against this idea: They are producing large quantities of everything else, for all markets. If Chinese investors prefer 1kg bars, then they are not competing for the 400oz bars. Doesn't explain why we don't see at least 1 bar show up.
8. [added Dec-2012] Chinese Refiners receive an export incentive to export silver, but not gold. (sent to me via email from Bron) This would explain the metal ratio discrepancy - i.e. why we see lots of Chinese silver bars show up compared with gold; Against this idea: still doesn't account for the complete absence of gold bars.
9. [added Jan-2013] China is just one country among several whose output is not represented in the data. (suggested by Costata) In other words, China may not necessarily be unique in terms of the gold bars not showing up, and the focus on this particular country gives rise to bias. ... good suggestion and it's true - unfortunately would require a whole heap of other data to compare against. The only thing I can suggest against this idea is that this still doesn't account for all the other reasons why we ought to see at least one Chinese gold bar showing up - i.e. based on expected huge output volumes and Chinese merchants trying to undercut the global market, we should expect to see a few batches.
... and lots more great stuff in the comments below!
The second reason I wanted to show you Julius Baer summary is because it is one of the last few 'good' sources of data that we have left (and hence, not finding a Chinese Gold Bar in the current data, we're running out of data to check). Leaving the China topic briefly, I thought I would end with a graph I've been working on for a while - a way of charting the relevant 'quality' of a bar list input. This is my early chart design, which shows the comparative number of unique document issues projected against the comparative market share. The (compact) result is 4 distinct quadrants which determine how useful the data will be for analysis. Funds with both Silver and Gold have been normalized into a single expression of market share, and the scale on both axes is logarithmic (i.e. the graph is designed to highlight position only). This chart has a very specific bias - it only looks at the bar list documents as related to data warehousing utility.
|How well do you know the transparency of your chosen ETF?|
Positions on the chart are approximate, for illustrative purposes.
Anything in the 'Whatever' quadrant has a very low market share, and very few unique documents being issued. We are not interested in these so much because they won't show a high resolution of change over time and not much inventory anyway. Goldmoney have now made their bar lists largely inaccessible so the point comes to figure out is it worth spending so much time chasing their data?
The 'Frustrating' quadrant is anything with a decently-high market share, but very few documents being created. It's no surprise that funds in this quadrant also have data issues - PSLV has incomplete data and the documents from ZKB do not list the Refiner Brand!!! (we can probably figure it out by matching the bar serial number patterns but it is very inprecise). For the most part these bar lists may otherwise have a lot of great and wonderful data, but we can't really put them to good use; hence, frustrating!
Anything in the 'Dependable' quadrant has a relatively high rate of document issue, but a low market share. We like these because even though they are small they give us a high resolution for changes, and generally speaking their attention to detail is HIGH - anything in this quadrant is generally reliable data and set an example for the other funds to aspire to.
The 'Brilliant' quadrant has the best of both worlds - huge inventory and lots of unique source documents being issued on a regular basis. It is no wonder these are the market leaders because they are well organized, and they are the most transparent because of the sheer volume of data on offer. Most of our studies from now on will focus solely on GLD and SLV exclusively because the other bar list document sources are simply not that significant by comparison.
Finally, there also don't appear to be any Vault Jumpers between Julius Baer and any of the other funds we have on record. This is odd, given the size of their Silver repository so I assume some kind of differences in the way they keep their data (I'll keep looking into this). There's only one last 'Brilliant' source of ETF bars data left to process in the great hunt for Chinese bars; the spreadsheets from ETF securities.
p.s. The graph above is evidence at how well Eric Sprott's social media advertising has worked - the bar lists backing his funds are among the worst in this categorization (my bias in this measurement already declared).
p.p.s. My praise for the big ETF's is not indicative of any specific endorsement; this is purely and simply a data warehousing exercise. Any anti-SLV and anti-GLD folk should be glad that so much good data is available since it gives us that much more rope to hang them with solid evidence for a level 1, 2 or 3 argument.
Updated: 20 November 2012 - to better demonstrate the 'Geography Isolation' effect, I thought it might be worth showing what the database has for Sprott's GOLD. Similar to the analysis on Sprott's Silver, he seems to have sourced the bulk of his gold from the Americas. Again, this appears to be an expression of cost of delivery and local stocks. Apologies I don't have the time to draw this up in a pretty graph - here's the raw output from PHYS:
|SPROTT's PHYS (latest) - 400oz Bar Refiner Origins by Country|
Update: 22 November 2012 - Dominic Frisby has just produced a more detailed writeup on the Hong Kong / China Gold relationship - it's well worth a read, in the context of what has been discussed here and also the research being done by Dominic, and of course to have a look at those charts that I mention in the comments (specifically, the 2010-2012 volume changes). His headline: 'Chinese Demand Could Send Gold Price Soaring'. Context: the charts originate from Nick Laird at www.sharelynx.com; the original data used for the charts is from this location http://www.censtatd.gov.hk/hkstat/sub/so230.jsp Regards, Warren