The Relative Sizes of the LBM OTC Market and the COMEX/GLOBEX Futures Exchange

Is the COMEX/GLOBEX ‘the be all and end all’ of the world’s precious metals market places, or does the LBM also play a large role, one that even surpasses the former, not only in volume but perhaps also in significance when it comes to certain aspects?

For me, this issue of the relative size, and therefore potentially the importance, of the LBM (London Bullion Market – see end note) in comparison to the COMEX recently arose in a small circle of those I regularly correspond with on matters pertaining to gold and silver. I sent around a posting by Victor the Cleaner on one of the more popular PM sites, in which he said that the LBMA dwarfs the size of the COMEX. One of my correspondents thought that was interesting enough to ask Ted Butler personally if that were true. Ted replied that he didn’t think that was correct. He has now stated this in his newsletter as follows (as ‘stolen’ by Ed Steer in his March 22 Gold and Silver Daily):

"For more than 30 years, I have noticed that on unique U.S. holidays, when Europe is open for business and the US closed, worldwide trading almost stops altogether in gold and silver. That wouldn’t seem to be the case if the COMEX were [smaller than the LBMA]. Further, the most dominant COMEX traders are also the most dominant traders on the LBMA and the OTC markets...and the allegations of manipulation are principally aimed at these traders anyway." ...

I don’t know why Butler has concluded that ‘trading almost stops altogether’ when the LMBA is open and the COMEX is closed for the day, unless he is just going on the volatility in prices on those days, which does not say anything at all about the volumes traded Over the Counter on the LBM, and could indeed be just because the US players are not active in either the LMBA or GLOBEX on those days, pushing apparent prices around. Whether or not the dominant COMEX traders are also the dominant traders on the LMBA is possible, but I would like to see some hard information backing up that assertion. In any event, how prices are established on the LBM, and on the COMEX/GLOBEX will need to be a matter for another post on another day if there seems to be enough interest in this effort, which begins the investigation of the relative importance of the two by establishing the relative magnitude of trades on each one. What would seem to matter most are those trades that result in the physical transfer of precious metals in the two very different ‘markets’, so that will be my focus in what follows, to the extent that relevant data can be readily discovered using only the Internet in a very finite amount of time.

The LBMA publishes monthly LBM clearing statistics for both gold and silver going back many years (, in millions of ounces transferred per day. The reported total value in US$ billions and the number of transfers also appear to be daily figures. The table below shows data for 2011, and although not explicitly stated on the page with the statistics, the associated clearing background page ( implies that the figures include both allocated and unallocated transfers. In the final three columns, I have added the number of business days in each month (including any holidays as I don't know what they were in the UK) and on that basis have calculated the approximate total ounces transferred per month, in millions. Quarter 1 totals are included for comparison with the second table, below:

Table 1: BD = Business Days
The yearly totals are 5,811,200,000 oz Au (180,730 Tons) and 45,403,000,000 oz Ag (1,412,048 Tons). Those figures are huge multiples of annual world production and, in the case of silver in particular, existing world stocks.

Fortunately, there are LMBA Liquidity Survey Figures for Quarter 1, 2011 ( - thanks to Bron Suchecki for the link), which help greatly in reducing the figures in the first table to more meaningful amounts.

First of all, the figures were assembled on a 'volunteer' basis', and only 36 out of 56 full members of the LBMA chose to participate, although they included all of the 'market makers'. Figures were requested for both sales and purchases with both other members and other parties. Separate figures were also tabulated for spot versus forward contracts and 'other' things like swaps and options, etc, but excluding deposits and loans.
Table 2: LBMA Turnover for 2011

* Data Source: LBMA, Comprised of data from 36 LBMA Members, including all spot and forward Market Makers, for spot and forward Loco London transactions
** The original LBMA document uses the dollar for currency symbol, so we assume this is USD.
*** Table split into two sections for clarity. For original table, refer source document

Points to consider:

From the text: 'It should be noted that the figures provided for trade between members were divided by two in order to avoid double counting. This is rather conservative in that many of the trades reported with members would be with members that were not themselves reporting .' This further dilutes the reported figures, perhaps significantly.

Since the turnover figures above are for 'transfers' (presumably meaning traded contracts) they include both turnover and clearing contracts, which were stated to occur at approximately a 10:1 ratio. Figure 2 in the document, a pie chart, seems to approximately confirm that ratio. So, perhaps the decimal points above need to be moved one place to the left to indicate cleared contracts, 90% of those being spot contracts.

The reported average daily trading figure of 173,713,000, multiplied by the assumed 64 trading days in the quarter, equals 11,117,632,000, close enough to the reported 10,943,926,000, so let’s say that the survey indicates that 11 billion ounces were traded in Q1. So, if we move the decimal point one place to the left and take 90% of that, we get close to a nice round 1 billion ounces of Au changing hands on the LMBA OTC market in QI 2011.

Since that survey number is known to be under reported to one degree or another, it is a fair ball park match for the 1.2 billion ounces in Q1 2011 from the LMBA clearing statistics in the first table.

However, figures from both sources would seem to include both unallocated and allocated metal. If we then take Jeffrey Christian's figure of 100:1 for paper versus physical at face value, the decimal points would need to be moved another two places to the left to produce figures for allocated physical gold, producing 12 million ounces settled in Q1 2011.

On that admittedly rough basis, a total of 58 million ounces (1,804 tons Au) of real physical gold may have actually changed hands in London for the whole year, taking the clearing statistics in the first table,at face value and then moving the decimal point three places to the left.

For silver, also from the first table, the corresponding number would be roughly 454 million ounces (14,119 tons Ag) of physical metal changing hands on the LBM in all of 2011.

What about the COMEX, then?

As the final figures above are an attempt to approximate the number of physical ounces of Au and Ag that change hands on the LBMA OTC market on a yearly basis, to make a comparison with the COMEX I needed to come up with corresponding figures. The similar figures most readily available to me are what Harvey Organ provides in his Daily Gold & Silver Report ( Around the last trading day of the month, he provides the total number of contracts delivered and still deliverable, adjusting daily for the number of contracts that appear to be cash settled. In looking for the final monthly figures for 2011, my browser would only let me see the last day in his reporting weeks before the last day of active trading for a month, so the following figures may understate the total number of settled contracts for the month, but the remaining days usually only involve small numbers of contracts of which a number are often cash settled. So I am using what I could find easily. The figures are as follows, the dates being the day of the month of Harvey's reports that I used:

Table 3: Ounces and Tonnes Traded on Comex (source: Harvey Organ Blogspot)

On the basis of all these admittedly approximate figures for 2011, it does indeed seem as though the annual amount of physical gold changing hands on the LMB is roughly 9 times greater than on the COMEX, whereas for silver, the ratio is nearly 12:1.

So Victor the Cleaner was right about the LMBA physical volume being around ten times the COMEX in terms of physically settled contracts, and since Ted Butler thought this was just not the case, one might conclude that he is simply fixated on the COMEX and doesn't appreciate what is going on in the rest of the world. However, before one could legitimately reach that conclusion, there are other things to take into consideration, which will be covered in a subsequent posting or two.

NB. The London Bullion Market Association (LBMA) operates the London Bullion Market (LBM), but the two acronyms are often used interchangeably to refer to either or both. It is important to understand that the LBM is NOT an exchange where prices are ‘discovered’ and disseminated. It is an Over the Counter (OTC) market where buyers and sellers privately agree on the price of a given transaction, much as you might haggle with your local coin dealer over an eventually agreed price you will pay for a particular ASE at a given time. Just as the dealer might agree to a better price if you promise not to tell anyone else what it was, it is in the interests of the participants in the LBM not to advertise what particular prices they settle on.


Kid Dynamite said...

and herein lies the biggest problem with the Metals-sphere:

major "trusted" sources like Ted Butler have no comprehension of the markets they are trying to analyze. I don't care if Butler has been analyzing silver for 30 years: it's clear that he doesn't understand how ETFs work (from his misguided confusion about short selling), and most others clearly don't understand how futures exchanges like the COMEX work.

Then there's the "dark" market LBMA, which dwarfs these visible sources.

So we have futures and ETFs, which provide visible data which is still not understood by the "analysts," and then the LBMA, which is bigger, and more opaque... In other words, visible data which is misinterpreted and misunderstood, and invisible data which is simply ignored.

and we wonder why the calls for Armageddon never come?

The Big Setup said...

"and we wonder why the calls for Armageddon never come?"

...Of Course, because everything LASTS FOREVER..

Kid Dynamite said...

no - Big Setup - it's because the people calling for Armageddon have no idea how the markets they are trying to "analyze" actually work.

Harvey Organ, for example, is no more capable of analyzing the metals than I am of fabricating drugs as a pharmacist - which I believe is his natural calling.

ps - Hey Screwtape files - what's the story behind the little rodent logos for each contributor?

The Big Setup said...

first of all, we have no idea if any figures that that are supplied to us are lies, real or just made up. The government and its entities lie from pills to shots to unemployment to debt levels to Fort Knox gold. I can't believe ANYTHING that is spouted from those mouths. We are just tiny tadpoles in an ocean filled with pollution.

Anonymous said...

[shocked and appalled] Warren - the nasty man called us... sob... "RODENTS"..!

@KD - you may know everything there is too know about ETFs and ETNs but you, sir, are no zoologist. ;-)

The 'about Screwtape' section tells you everything you need to know. We are a collection of five prosimians. What could possibly be more traditional for a PM blog?

Welcome aboard SLL! An outstanding first article. Looking forward to the next!

Bron said...

This analysis looks at comparing physical trading. Because the size issue is more about price discovery (or manipulation if you will), which seems to be the focus of Ted Butler, I tend to favour just looking at total turnover or dollar value traded. Yes this includes paper and physical trading, but what I'm looking at here is the "weight of money" moving in these markets.

My approach is summarised in this comment on Gene Arensberg's Got Gold Report :

The LBMA survey "...for Q1 2011 reported net gold clearing of 1,183 million ounces vs turnover of 10,943 moz, a ratio of 9.25:1. It was only gold, but I suppose we could apply a similar ratio to silver net clearing figures, which would make your 197moz into 1,823moz of trading (per day). ... CME reported 1478663 contract volume for Feb 2012 which at 5000oz per contract divided by 21 trading days = 352moz. Compared to my rough 1823moz figure, this make London OTC market 5 times larger. So no, COMEX does not have by far the most influence on short term price discovery."

Applying 9.25:1 to Larry's 8672.7moz silver clearing for Q1 I would get estimate of 80222.475moz of turnover on LBMA. From the CME website (not Harvey) I get reported volume for Q1 2011 silver futures of 4790094 contracts x 5000oz per contract = 23950.47moz. On that basis LBMA is 3.35 times larger.

Whether the figure is 9, 12, 5 or 3 times larger, I think no matter how you look at the numbers, LBMA is bigger than COMEX.

Anonymous said...


good to see that you found such a reputable home for your article. I also say sorry for not responding earlier. There is something I wasn't happy about, and Bron has basically already mentioned it.

I don't think one should compare physical trading. This is because of the following fact: If I were a wealthy individual, a family office, or a hedge fund, and I wanted to buy allocated gold or silver, I would never ever think of buying it by going long at the COMEX and then taking delivery. I would just phone JP Morgan, Scotia Mocatta, or HSBC, and simply order the stuff.

COMEX is not meant to be a venue for physical trading. It is also not meant for taking large deliveries. In the silver market, for example, the industrial users typically have direct contracts with the refiners that are for a quarter or for a full year. None of them buys at the COMEX by going long and then taking delivery.

The delivery mechanism at the COMEX is just there in order to make arbitrage possible. Once it is possible, the prices of the near futures will match the spot price (more or less because there are still different locos involved) plus contango.

When you see large deliveries at the COMEX, this just indicates that during the previous month, you had a directional 'fast market' in which the arbitrageurs between the OTC market and the COMEX were active. In some sense, COMEX deliveries are a sort of friction that would not happen in an ideally efficient market.

So I agree with Bron that one should compare the US$ volume traded. In addition to Bron's figures I might add that there used to be a rule of thumb that the OTC market is 3-4 times bigger than reported in LBMA statistics. I.e. when LBMA says they clear 600 tons of gold per day, that indicates an OTC volume of 2000 tons per trading day. That makes OTC roughly 10 times bigger than COMEX if I remember correctly.


Bron said...


That 3-4 times rule of thumb I think came from some presentation at a LBMA conference and was an (informed) speculation by the speaker (don't have the link to hand).

That was pre the liquidity survey which now shows it is, at least for Q1 2011, 9.25:1.

Slow Loris Larry said...

Now that comments on my posting seen to have run their course, I would like to especially thank Bron and Victor the Cleaner for their helpful additions to my post, which corroborate and expand my basic argument that the LBM is a much larger PM market than the COMEX (which was Victor's original assertion that Butler took issue with). Bron's point that the overall dollar volume of all trades is more important than the amount of physical metal that appears to be changing hands on the LBM is valid (of which more in a later post), as is Victor's complementary point that the amount of COMEX contracts standing for delivery is not really very salient as there are much better ways to obtain the metals in volume, even in North America. My reasons for limiting my quick little research project to transactions that involved (or at least could have involved) physical delivery of metal was 1) because so much attention is given to the 'let's get physical' aspect of the markets in the blogosphere, and 2) that the immense volumes of paper transactions on the LBM need to be connected somehow to something real if any sense is to be made of them (at least superficially, by me anyway).

I would also like to thank Warren James for all the help he gave me, including formatting the tables in the article in html and walking me through the hoops needed to actually get something posted under my new little prosimian persona.

Anonymous said...


I hope you are not shy and whenever the usual 'argument' of Ted Butler comes up, you just point people here. Everyone who is interested ought to have a chance to find out.