A different view of GLD (part 2 of 4) updated

This is part 2 of a perspective on GLD's inventory you won't find anywhere else on the internet; an edutainment series using the ETF bar list data as a base. In Part 1 we looked at what occurs when we look at a specific set of inventory, here is the updated graph (using the data from July 2013). This time I have added some extra data points to help accentuate that the line indicates the measurement between two chosen points only. The additional measurements show bullion for a shorter 2-year range.

The last few months of removals has started cutting into GLD stock originally accumulated more than two years ago (i.e. this is shown where the corner meets the inventory curve), whereas that wasn't the case in February. For our 'overall' measurement, the dark yellow line has moved DOWN (from dotted line) since last month, so we can say that over 1,000,000 ounces of gold removed in the last month, was taken from stock present prior to June 2011. But while this is interesting, it is no particular conspiracy as it just reflects the fact a lot of gold is being de-listed.

Also bear in mind this chart summarizes a lot of detail. Don't read into the figures too much.

Speaking of summaries, here's an updated info-graphic showing GLD activity for July. There was an 'add event' in the bar list data (with most of those 'new' bars originating from ETF Securities stock) but the addition can only be detected from the bar list figures - if you're only looking at the trade settlement spreadsheet (link) then you won't see an increase in GLD inventory for July.

I'm still experimenting with the layout and information content. Feedback welcome.
Let's focus today on gold 'removal' which everyone is talking about. As you will note from above, with the bar list we have access to finer inventory detail compared with other analysts. Let's use this level of detail to analyze Turd Ferguson's comments over 'same amount withdrawals' (link):
"... Anyway, I call bullshit. To be consistently redeeming the exact same or multiples of the same number of tendered shares is statistically impossible, akin to getting struck by lightning repeatedly." - Turd Ferguson (TFMetalsReport.com)
His angst on this particular issue started with three particular withdrawals each for approximately 1.5 tonne (on TFMetalsReport here and here), with another similar just recently in the first week of August. I have no knowledge of 'tendered shares', and as such I'm not going to tackle his carefully crafted statement. What I am interested in however, is the propensity for 'redemptions' to occur in tonne or half-tonne blocks. To begin deconstructing this, let's look at the data:

Bar List data Changes for July 2013 (added/removed)
Entries highlighted yellow are interesting - The 216 bars removed on the 24th are the same bars which were added on the 15th (confirmed from database). GLD's inventory increased briefly by about 2.7 tonnes using gold from another fund (ETF Securities) and then the very same stack disappeared 9 days later. Weird right? I am no warehousing expert but to me it does suggest these bars are 'the pallets currently being moved around'. You'll note even in this data there are many round numbers, with Turd's 1.5 tonne removals (highlighted with a black border). You can download the exact comparisons from this spreadsheet (link).

Is it common for redemptions to use exact amounts? Using the database I have assembled a rather crude copy of Victor's Histogram, which highlights (in blue) the removals mentioned by Turd. I hope this demonstrates that statistically, when charted against the backdrop of their fellow data points, these seemingly suspicious amounts actually have nothing remarkable about them.

Chart showing number of bars added/removed between mid-2011 and mid-2013.
Some limit of error reading does exist since the data needs a bit of cleaning up.
[Total source table size is over 65,000,000 rows.]

Save your breath on threatening me ... I'm after a much bigger prize and I did promise 'a different view'. One of the things which stands out in the data is that removals are often from arbitrary sections of the bar list. A single day's removal events are often constructed of multiple contiguous blocks in the bar list data. We know this because our database records the line number and page number - we also know for a fact the data is based on Excel, so removals are basically the removal of rows inside a spreadsheet. Treating each individual segment as a separate 'bullion removal' event, changes the graph!

A different view of GLD! Pallet based removals!
Showing all the events side-by-side. Interesting that the very large removal, was
still one continuous block in the spreadsheet data. Many entries are multiples of 80.

The 1.5 tonne items show up in the data as separate bullion blocks, I've left them highlighted blue so that you can decide for yourself whether they are still noteworthy. As you can see, blocks of 80 and 40 bars are incredibly common in this view, that is 1 pallet and ½ pallet respectively (and 80+40 bars = approx 1½ tonnes). Bron Suchecki has forwarded me an explanation as to how this would occur in relation to vault operations and I'm quite happy with it (the explanation) **; later in part 3 we will chart these individual 'redemption blocks' as our 'bars selection of interest', revealing their own unique stories.

UPDATED 13th August: I've had feedback about whether the charts above are histograms or not. They aren't, although they look like it and was inferred. The ambiguity was meant to reflect the lack of precision available, for example the very large 'contiguous blocks' don't show us pallet-level information (though the smaller ones do). The y-axis is 'number of bars', the x-axis represents an individual event, with the highest amounts on the left, smallest amounts on the right. The graph is meant to demonstrate as a quick visual only. Here's a bit more clarity on the last graph, you can see that many of the events come in multiples of 80 ... e.g. there are 14 individual events of exactly 160 bars (2 pallets).

Close up, showing example of multiples of 80
Regarding basket sizes, most of the redemptions are always clean basket sizes except for twice a month where it is uneven (normally around the 10th or the 20th of each month), the current theory is that these events represent the fund selling bars as operating costs. The basket-based analysis could definitely have an article all of its own. We also need to establish (somewhere, somehow) the relationship between baskets and pallet movements. For example, here is the individual breakdown of Turd's 5-basket redemptions. Do we take this to mean that some of the basket redemptions were 2+3=5 because of the ratios? Some detail is lost in the process, but the individual pallet movements are unmistakeable.

Actual database output isolating the Turd entries.
Does the separate half-tonne pallets indicate these
were separate customers/recipients/participants? Not sure.
Until next time,

Warren James
Screwtape Files

NOTES: These graphs DO have limit of reading errors, approximately ± 3% because some 'changed signatures' are still showing in the older data as removals. The 'contiguous blocks' analysis (the last graph) is a brand new analysis set, so it's still really primitive and I expect to gain better detail over time. This is also subject to it's own limit of reading error if they do weird things with the spreadsheet records (e.g. shuffling entries). The 'contiguous blocks' chart does not currently incorporate any dark bullion measurements.

We welcome criticism, in particular pointing out any flaws with regards to analysis (because it helps refine it), just be polite.

** Updated 12th August --> See Bron's fine comment below as a great example of how the 1 tonne batches arise in a custodian environment. This is a refined version of the information he shared originally with me.

Bar List Random News:

On June 27th, the bar list for GLD was missing 871 pages - similar to the October 2009 anomaly noticed by Rob Kirby* but this time it went completely unnoticed by the metalsphere. I analyzed the missing bars but couldn't find a pattern, the header repeats itself on page 294 so it looks to me just like a 'print-to-pdf' error. The mistake was not present in the bar list the following day.


Slow Loris Larry said...


Another tremendous effort. Keep on cleaning up the 'bullshit'. Can't wait for Part 3.


Anonymous said...

The reason you sometimes see the same repeated amounts added/redeemed is that it is done in basket amounts. 1.5 tonnes is 5 baskets.

Beer Holiday said...

I think that in general warehouse data, you expect the changes in item numbers be multiples of some fundamental quantities such as

- EOQ Economic order quantity
- Minimum sales quantity
- Pallet/carton/box number of items.

Especially at the small-number-of-item-changes end of the spectrum.

I don't know about PM vaulting, however. In GLD is number of baskets such a quantity as Daedalus (?) points out above?

Also cubic and square numbers show up a lot, because they are the natural packing volume numbers of box shaped items. I'm not saying I think that's the case with recurring 216 bar movements but...

216 is 6 cubed. So if you want to visualize that gold, it's a cube 6*6*6

666 - Does this mean it's the devil who is draining the GLD haha :-)

Beer Holiday said...

I got that a bit wrong, you can't make a cube of 6x6x6, (it was mostly an excuse to make a "the devil runs GLD" joke. )

Warren James said...

@Beer Holiday, yes, the basket observation (by 8f6d4ec6 above) basically nullifies Turd's hyperbole without the need to identify the underlying bullion segments - I wish I had thought of it although to be honest my articles don't generally start out with 'debunking' in mind (by the time I've finished the data results it's often at odds with some other narratives you'll find on the internets).

One GLD basket size is 9659.70 oz, and it's the reason why a lot of the tonnage shows up in multiples of 0.3 tonnes. Re: sizing, you'll find most pallets from the Bob Pisani video are 80 bars to a pallet, with some incomplete pallets as well. For what it's worth, the assumption that bars from a single refiner will 'stick together' gives us another avenue to determine pallet boundaries in the data as well, but it will take some time to tease apart the patterns.

I'm basically hoping we'll be able to extract some stories about what is happening with the gold, before it disappears completely from public view.

Bron Suchecki said...

Excellent analysis. In the wholesale markets gold is packed 1 tonne to a pallet. That is its standard lot. Just like when you go into the supermarket and see people filling the shelves the food comes in cartons, of say 12.

Looking at ETF movements is like observing data on supermarkets orders from suppliers. So observing that supermarkets seem to order toothpaste in lots of 12 and noting that individual sales of toothpaste can't always round up to 12 is not an OMG moment.

Note that for any ETF the custodian is also an authorised participant. And note with GLD the APs give unallocated gold to the custodian and it is the custodian who has to find the physical.

As custodian they want to deal in 1 tonne lots where possible so they don't have different owners of bars on the same pallet - easier to manage from a physical redemption point of view as you don't have to pull individual bars off a pallet if the owner wants the physical delivered.

So lets consider the following transactions for a day:

AP#1 delivering 1000.345 oz
AP#2 redeeming 534.467 oz
AP#3 redeeming 2500.000 oz
Custodian as AP#4 redeeming 134.467 oz

Custodian is faced with net redemption of 2168.589 oz. They identify 4 x 400oz equalling 2135.458 oz, so they change their AP#4 redeeming order to 101.336 oz and book a 33.131 oz long position to their trading account (they hold 33.131oz worth of GLD shares).

Next day they will throw the 33.131oz position into the AP transactions and net/balance everything all over again. That is how ETF redemption can consistently appear as 1 or half tonne lots.

So just as the supermarket sits on a bit of stock and varies it so that they just order toothpaste in cartons of 12, so does the custodian acting as an AP sit on some GLD shares to smooth its physical redemption/creation to make its vault management easier.

Bron Suchecki said...

"sits on a bit of stock and varies it so"

meant "sits on a bit of stock and varies its orders so"

That is, a retailer will round their actual sales up or down to lots of 12 and order that amount.

S Roche said...

Fact based research and sanity in the metalsphere - hallelujah! Great job.

One correction: you mis-spelled Turd's hyper-bowl... sry.

Kid Dynamite said...

Bron -
I fear your comment will confuse some observers as it deals in hypothetical basket sizes...

Warren -
I think your histogram would be better served by charting "# of baskets redeemed" instead of number of bars.... I'm not sure, but I think your comment in this thread says that's what's coming in the future. # of bars is misleading, as basket size changes over time, and the same basket size can be made up of different bars.

aside, it may be just me, but I was unable to view the x-axis on your charts: what are the most common amounts near the peak of the curve?

final question, Warren: which ETF securities product did the bars come from? where is that ETF Securities gold domiciled, and when did it leave ETF Securities?

Warren James said...

Hi Kid Dynamite,

Quite true - the focus of this series has always been on number of bars by default because that's the atomic unit we track in the database. To be honest I had momentarily forgotten about basket-based redemptions, and I'll work hard to providing in the future, different measurements.

Choice of units of measurement is really the issue here - Turd looking at quantity of shares redeemed. Victor usually looks at total ounces removed using spreadsheet as source, I can see blocks of gold at the pallet level at the vault level, and measurement in baskets make really good sense when talking about ETF mechanics. I was wrong to ignore basket sizes but I will do my best to factor them into future comparisons. I've updated the spreadsheet (link) if anyone is interested in the lone data points making the last histogram - apologies for the rawness.

p.s. The x-axis of those last two simply represent the number of 'events' in the data - so there's about 169 removal days for the last few years, and if we split those into individual bullion blocks we get over 500 mini-events.

With regards to those ETF Securities bars, the product they originally came from was Gold Bullion Securities Limited, the last record we have is end of Feb-2013 before they showed up in GLD ... but the disappearance point could have been anywhere between Feb and June 2013 as there is a gap in our ETF securities data (where they changed their spreadsheet format).

ssgtrader said...

hi Warren

I read a lot about shortage of physical gold in goldbug blog, in you mind or from your source is it true ? Thanks !

Warren James said...

@ssgtrader, great question. I have been searching for over two years for evidence of 'metals shortage' and I have found none -> in fact the data points to an abundant supply ... i.e. it appears you can buy as much gold as you want at market prices with no specific premium. Whoever is phrasing the question needs to qualify whether they are talking about the wholesale or retail market, as they are two different beasts - often confused as the same thing.

Having said that, I am intrigued with the GLD drain which is occuring ... there seems to be a lot of change of ownership going on.

In this article I forgot to link back to the main 'Bullion Bars Project' -> there's lots of articles there which talk about gold production that we've observed through the database. Basically even simple artifacts like the Johnson Matthey Sequence (the sequential run of 6-digit serial numbers) indicate that the refiners are happily creating big bars of gold (and we can approximately trace these by date). If there is a shortage of gold, perhaps no one has told the refineries about this!!!

ssgtrader said...

Thansk Warren, that is what I suspect all along but the rattle about shortage is getting to feverish pitch as the price goes lower. As I know gold is rarely "consumed" so whatever was dug up is always there. The only problem is, with the crash in price will those "available" in the market sudden being snapped up and made "unavailable", especially with your article on draw of GLD inventory.

ssgtrader said...

Just to add, there are also rumours that GS added US$5b long positions and JPM turning long on Gold. Althought from COT commercials are still net short.

Warren James said...

I guess it depends on perspective - there's still 29 million ounces sitting in a very visible space. To get them all you have to do is outbid current GLD holders ... technically anyone can take part in the redemption of shares for gold ... I guess from a proof perspective the biggest problem is we don't have anecdotal information from anyone who has gone shopping for a billion dollars worth of gold. Yet even that kind of purchase would only make a small dent in GLD .. I simply can't translate this to 'shortage'.

I am on a mission to find out the finer details about the GLD drain - for my own part we'll have a few more clues from the bars data once I've completed Parts 3 and 4. hth.

S Roche said...


There is a shortage or retail (manufactured) bars and coins for immediate delivery in some areas of high demand. There is a delay. Such delays cause retail buyers concern, which leads to suspicion and distrust. Understandably.

Yesterday's confirmation of recent record outflows of gold from the UK, largely to the refineries in Switzerland in order to be manufactured into bars and coins to meet retail demand, (and possibly for storage there), reveals that the wholesale market (400 oz bars) is able to meet demand for delivery at the reduced prices of the past few months.

We will know when the wholesale market can no longer meet demand for physical gold delivery at these prices, the sign will be that the price of gold rises further...

Those who promote and distort shortages in precious metals have an agenda to sell manufactured precious metal or a condition, known colloquially as bait-clicking. That they see themselves on a mission of enlightenment says more about them than their mission.

S Roche said...

Or another way to put it...

Despite the record demand for gold recently buyers were unable to keep up with the record(er) supply of physical gold in the wholesale market and prices plunged.

So, there is actually an abundance of gold, not a shortage. Which, on the basis that nobody knows anything, sounds about right to me.

Gary Morgan said...

'Despite the record demand for gold recently buyers were unable to keep up with the record(er) supply of physical gold in the wholesale market and prices plunged.'

Interesting words SR, but can you back it up with any facts, for example, this record supply of physical gold? Any proof? I've been reading about mines going bust, ramping down production, and scrap supply dwindling....so where's this record supply coming from?

Also, if I went to Bron's Perth Mint and bought a few tonnes of physical tomorrow, you think that would move the (paper) price at all?

S Roche said...


GLD, (as one example of Western dis-hoarding of investment gold), from 1,300+ tonnes in 2012 to 900+ today leading to Macquarie Bank via Reuters 2 days ago reporting record exports of physical gold from UK to Switzerland, 240 tonnes in May alone:


The significant stocks of gold in London appear to have supplemented the flow East.

You would be aware that the price dropped from 2012 to June this year, coinciding with the availability of this physical gold. That price action shows to me that the demand at the higher prices was overwhelmed by the supply and then that reversed at the low (so far) of $1180.

I doubt a few tonnes would move the spot price at all, as that is well within usual dealings in the wholesale market as I understand it. I am not a participant. Your point presumably is what amount of physical gold, purchased in one lot, would it take to move the market? If you would care to study the figures over the last 8 months you would be able to answer your own question as the market has seen considerable sales of, and delivery of, physical gold and the price has responded in both directions.

S Roche said...

@ Gary

This seems to cover your question and my answer more fully:


Gary Morgan said...

SR, thanks for your response.

It appears you are putting the chicken before the egg.
Demand in the East shot up after the paper price of gold fell, and has more than kept up with the sales from GLD, as we've all seen the reports of 4-6 delays for gold buyers in China and Dubai.

I also think you are mistaken regarding physical sales/purchases of size moving the market. There is only one gold price, and it is a paper price. The central bank in India recently confirmed a 92x higher level of paper trading versus physical.

S Roche said...

Oh Gary,

Please don't ever ask me a question again.

Your mind is made up. Good luck with that.

Gary Morgan said...

Oh SR, my questions were merely designed to highlight that you have the closed mind.

Your comment 'We will know when the wholesale market can no longer meet demand for physical gold delivery at these prices, the sign will be that the price of gold rises further...'

...shows how little you grasp the geopolitical aspects of the 'gold' market. But you are just a trader of paper I believe, so good luck to you too.

ssgtrader said...

hi SR,

looks like Goldman is really buying quie a lot of GLD, it is not rumours.


ssgtrader said...

Contrary to my own believe that tapering is going to be bad for PM, I am going to start probing on the long side. There must be something they know that we are still unaware of.