GLD Inventory Large Addition Analysis

Just like the gold price itself, the inventory of GLD was incredibly boring for the last few months then it got some excitement all in a hurry with 668 gold bars being added to the inventory on the 30th May bar list (shown three days earlier on the trade settlement spreadsheet). That's a big increase for one day, let's take a look at the bars themselves.

I was expecting a large percentage of the add event to be from gold bars previously seen in the inventory (dark bullion). I wasn't disappointed. 487 bars were previously dark ... i.e. 23% new, 73% old.

The 181 bars which were 'new' were branded Johnson Matthey -> this is no surprise because the vault generally contains up to 25% JM bars, but also worth noting, none of the serial numbers on these bars are in the high end of the Johnson Matthey Sequence which means they are actually 'old' bars (in our classification index: " new cold bullion"), or in other words the 'new' gold did not come from a shiny new delivery from Johnson Matthey > it came from existing stock.

Similar to other recent additions, the 487 bars which were previously dark also have a dark history. Just to recap, a 'dark history' event is recorded for a bar if it disappears and reappears on the bar list at a later time - the current record is 4 times for GLD stock. You can see below, the speckled history of this bullion.


Unlike previous additions, most of this resurfacing gold has been in the system for some time. 22 bars have been missing for over 4 years, first sighted in the earliest 2009 bar list (!) and over half of the bars from this addition, were first seen prior to 2012. The most recent activity was 80 bars which were last removed 3 days prior to the add event.

There is currently no sign of my predicted bars to return, but that's okay there's still another two months before that prediction is invalidated.


So those are just observations, what does it mean (if anything)? On the surface, the notion of 'regularly returning bullion' would seem to indicate some kind of 'ownership' BUT this seems unlikely, it is more just a statistic which you would expect similar to a maths question : 'x red and y blue marbles drawn are from the bag at random then x% are added back ...'. This is the most likely explanation since although the add event was quite large, we are still only talking a really tiny percentage of the GLD peak inventory which occurred on 27th December 2012 at 109,318 gold bars (current holdings 62,928).

What can be said is that the entire addition was from old stock, so I just want to point out in a random sample from [ Unknown Stock + 46,390 missing bars], there is still enough gold in the vault background to introduce 181 bars that we've never seen before, so the unknown stock levels (vault, london, etc) are a lot higher than most will give credit for.

Some readers may know that I'm attempting to establish some theory around 'how the fraud works'. I still think the whole modern financial sector is fraudulent by nature and I have no doubt that manipulation exists but I just don't think it exists in the manner described by the other silver and gold blogs. I offer the following suggestion:

Consider the last week's events from the perspective of one of the Authorized Participants. An entity had ownership of some 269,902 oz (gross) of gold and for whatever reason they wanted to create some GLD shares. We presume they sold these on, and at some point in the future they will buy some back. Answer the following question given the week's price action : Was this arrangement was good for (a) the Authorized Participant or (b) the holders of GLD shares (including the newly issued ones)? Regardless of the order of cause and effect in this situation, it's just another shining example of how these guys have the game stitched up nicely.

Be careful out there,
Warren

13 comments:

costata said...

Sleuthing with attitude! Good luck in your hunt for the nature of the scam(s) and thanks for the ongoing research.

Cheers Warren.

Marks said...

Can we conclude/theorize that new sequenced 400 ounce bars from JM and others are being sent direct to Switzerland to be converted to 99.99% Kilogram bars...?

Warren James said...

Hi Marks, that's a decent theory but there's no public data we can draw on to support it. But yes, we could create a map of when we see new JM bars being added to visible stock and use averages to infer volumes. The biggest problem is that the Gold ETF's only still represent a small sample of the stock in London -> we only get to see it when it passes by the GLD window.

Warren James said...

Thanks Costata. p.s. uUnderstanding the scam is a double-edged sword; once you figure out how it's done there is a large danger of joining it ;p

I look forward to the day when GLD inventories increase in a big way because it will give us even more data to work with. Personally I expect the GLD drain to halt and reverse sometime in the next 2-3 months. Let's see if I'm right ..

burningfiat said...

Hello Warren,

Nice work with the bar-tracking!

Why do you expect the GLD drain to halt and reverse sometime in the next 2-3 months?
Just a feeling or something deeper behind it?

/burningfiat

Warren James said...

Hi burningfiat,

It's just an idea at this stage, based on T.A. on the inventory chart itself (one might argue that the inventory level is already building a base) but it's also combined with the theory embedded in my last paragraph.

For the rest of this year I expect more inventory coming back in even as the price gets smacked down towards $800. But I think long-term (2-3 year timeframe), the inventory is headed south again in a big way and this will happen as the price soars. I think that is the scenario which would flummox the largest number of investors and observers.

BUT ... I recommend no one take me seriously on that prediction, since my theories are only in infancy and currently quite lacking - but ... that's what I currently think and is the thesis I'm setting out to prove/disprove.

burningfiat said...


For the rest of this year I expect more inventory coming back in even as the price gets smacked down towards $800


Bold theory, lovely! I would love to see $PoG at $800 and would certainly be flummoxed if that entailed rising GLD inventory at the same time.

Let's see whether it holds!

Bron Suchecki said...

The question I have is if you graphed the observed ETFs actions as a complex tree diagram http://www.mathsisfun.com/data/probability-events-conditional.html what we know is

a) number of blue marbles (bars that have been made public via an ETF)
b) number of blue marbles drawn back out (once being replaced back into the "bag")

Could a smart statistian work backwards to work out c) the number of red marbles (unpublic bars in vaults)?

Complicating it is the fact the drawing out and placing back is not necessarily random and also we don't know how many red or blue marbles a taken out of the bag forever (eg melted into kilobars).

Bron Suchecki said...

I suppose a statistian would say that if we assume a x% of marbles been taken out of the bar forever, then they could solve for # of red marbles, or vice versa?

Warren James said...

Hi Bron, re: statistics - yes I believe this data can reveal estimates about non-public bars in vaults, even allowing for 'dead bullion' (name courtesy of Marks). One of the keys would be the Johnson Matthey Sequence - on the assumption that large blocks tend to stay together. The good news is that the more data we collect, and the more volatility in gold, the better chance we have of pinning those estimates. Along the way we could establish some principles regarding gold bar locality and propensity to shift. I guess this is a long-term goal. ;)

Marks said...

FWIW, I am theorizing that most of the "Dead Bullion" is in fact via newly sequenced bars. Maybe some business reason such as reduced shipping cost that leads Swiss Refiners to choose newly minted 400 oz bars to melt into 99.99% kilogram bars for shipment to China. Or could it be that Swiss Refiners are not melting down 400 oz bars at all, i.e. they are using miner dore to melt direct to 99.99% bars, and thus bypassing the minting of 400 oz bars altogether...? (If so, this would imply less newly sequenced 400 oz bars).

Bron Suchecki said...

The commercial objective of any refiner is to avoid making any 400oz bars as they do not attract any premium. If a refiner has demand for kilobars or other value added product they will go straight from dore to that product (intermediate step from electrolitic cells to bars is granules, not 400oz bars).

It is possible that one refiner may not have any immediate kilobar or other demand so makes 400oz bars but then a few weeks later demand returns so a bullion banks sends those 400oz bars to a refiner for conversion into kilobars.

Warren James said...

It's worth pointing out that my theory is at odds with Trader Dan http://traderdannorcini.blogspot.com.au/2014/06/gld-holdings-continue-to-rise.html, so this should be an indicator that my theory is wrong.

Inferring a relationship between gold stock and price is a difficult thing because of the 60-year supply overhang http://goldchat.blogspot.com.au/2014/04/why-no-direct-relationship-between.html but it's still my view that the big guys use their big stash to help nudge price in their desired direction. After all they built the entire setup!! It's no stretch of the imagination to suggest there are ways to exploit the structure!