Quick Analysis on Sticky Gold (chart)

A very quick chart for Gold price (in USD/oz) to demonstrate the gold doldrums. For those who debate Gold's proper value : the market says it is fairly priced at around $1,200/oz. The graph below shows the sample of the mid-point price points for this year, expressed as a percentage of all price samples. The underlying data is for XAU_USD currency pair, from Oanda's trading platform, analysis of all M1 intervals available for calendar year 2015 to date.

$30 range around $1200/oz is highlighted in yellow.
Data points are for Jan - May 2015, from Oanda.

Defragmenting GLD

This article presents some recent bar list work and shines that spotlight even deeper into the darkness of the vault. Does anyone remember the old 'defrag disk' programs we had before solid-state-drives? The screens were beautiful to watch as they worked! Seeing the individual blocks of data move around gave a better idea of what our data was doing and showed a different side to the disk drive we only ever saw the outside of.

A typical defrag program - showing data allocations on a computer file system.
The software would remove blocks from one location and write them in another.
Some files did not move at all (and you could see which blocks stayed).

The mechanic shares some similarity with a gold vault, an analogy close enough that we can take advantage of - a specific fixed area filled with lots of small (identical) units which take up space and are easily moved around.

Sunday pre-game, 7/1/2012

UPDATE: I attacked Turd Ferguson in this post on the grounds that he disrespected a Screwtape colleague. I have since reason to believe (as per Turdite and commenter LeGrew's insight) that I might have jumped the gun and accused him unfairly, so I have deleted that part of this post, which, upon reflection, was uncharacteristically nasty, on account of my oxycontin withdrawal pangs at the time. 

Alright, now to the metals. We had a big move to close the week and month. These charts all speak for themselves, but I will speak a few words for them regardless (in captions):

Gold YTD. Watch the 40-day exponential moving average, which I follow in place of the less informative 20-day.

James Turk: Learn from the Master of Chart Analysis

I'm proud to announce another fantastic opportunity for Screwtape readers today. James Turk, the spokesman of GoldMoney, seems to be the go-to person for commentary on the future of the gold market. In particular, he is known for his forensic interpretations of charts, in which he bravely and with great integrity puts to one side his own business interest of selling physical gold in order to objectively make the case that readers and listeners should buy physical gold.

So, in order that our readers may learn to profit more fully from Turk's incisive technical analysis, I will set out here some of the basic techniques that he uses, with a view to showing how trade-able conclusions can be reached that will ultimately lead to enormous gains in your personal wealth.

SLV Database #2

(Guest Post by Warren)

A quick update from the SLV database. I was looking at the SLV inventory over the last two weeks of April. I cobbed together this chart (click for larger view) which compares the closing price of SLV vs. the inventory held according to the ounces published. It's a bad graph, but one of my first composite charts so be kind.



The graph is curious because on the 20th of May, the number of bars held takes a big drop and then peaks again on the 27th April with the high in silver (the gap in the middle is the easter holiday).

I have always wondered about the bars which disappear and then show up again. It stands to reason that because the 1000-oz bars are quite heavy, that movement in and out of the vault will probably be more like allocations rather than actual shipping (as highlighted by Bron).

So using the 13th April bar list as a reference point, I set out to find out how many bars showing up in 27th April list, were already present in the 13th April list, with a temporary absence of leave on the 20th April (in between).

After crunching the numbers in the bar serial numbers database, I have an answer - 8.3 percent of the amount added to the inventory between 20th April and 27th April. i.e. between those dates, some 9,954,185 oz of silver was added to SLV, and of that amount 1,197,039.8 oz (8.3 pct) was already sitting in the vault on the 13th April (14 days before). The table in the database for those date ranges is called "SLV_Analysis" so you can check my maths if you want - the connection information for the database is here.

I used this exercise to cut my teeth on the data processing and I give no personal guarantee about any numbers, this exercise was about 5 hours of work. I'm happy to share any details if anyone is interested (and again, the data is all there if you want to check).

What to conclude? Not much - approximately 1197 bars probably never left the vault and went 'dark' for a few weeks while SLV shuffled its data around, reappearing 14 days later. Either that or those bars left the vault and then got redelivered within a few weeks. I am personally more inclined to think that they have shitloads of silver - more than you can imagine - and they shuffle the allocation as best as they see fit to manage the price and futures, all that stuff, to make lots of money from hapless investors. Certainly it's more likely a change of ownership registration than a physical move, but it means that we've got more hard work ahead to identify these dark pools of silver. If anyone out there has any comments that can enhance my understanding, please let me know your thoughts, many thanks.

NOTE: This analysis indicates the SLV Serial Bars data is about right. If it were claiming all the bars were brand new, that would be unusual, and if all the inventory showed up a second time, that would indicate shenanigans, but a 8.3% seems about right.

[Update: 2012]

Since this analysis, we got a bit more sophisticated. For a full list of all future articles in this series, please check out http://screwtapefiles.blogspot.com/p/bullion-bars-database.html

SLV Database (Guest Post by Warren)

Time to test some claims again. Todays spotlight of scrutiny turns to SLV, the world’s largest stockpile of physical silver. For those of you looking for conspiracy, you will NOT find it here. What you will find, is analysis and questions of CLAIMS being made. The people running the trust CLAIM they have 366 million ounces of silver. Others CLAIM they don’t. Who is correct? And darn it, there’s money to be had depending on who is right or wrong.
Following GM Jenkins fine example of putting a chart to opinion, here is an anecdotal chart drawn from information available in the blogosphere. Note that each party has their own line of reasoning backing their stance (if I have unfairly presented anyone here, let me know):



First, why does it matter? Well first up – it’s a very big pile of silver. If the real silver is there as CLAIMED by the iShares Silver Trust, then most of the other stories we’ve been hearing don’t make sense. After all, it’s a simple case of whoever wants physical silver just buys a bunch of SLV shares and redeems them at spot price. No premium-laced cash settlements necessary. No 22% NAV premium required. No tightness of physical market. No delivery problems.
Now, I could write a book (or a very good action movie) about all the various counter claims – unless living under a rock then you are familiar with them already, and therefore I won’t discuss them in depth. Instead I want to focus on doing some hard concrete objective analysis. I also want to expose the methods for my research, to demonstrate that I’m not making any conjecture. Then I want to go to the extra step of making this accessible.
Here’s the gambit:

1. SLV publish their bar serial numbers in a new PDF every week. Very transparent.
2. There’s a lot of (public) information there, but it’s difficult to study because of the format.
3. Let’s collate this information so that we can analyse it better (a relational database).
4. Let’s put the database online and open it to the public so that anyone can access the info.
5. Let’s graph this data and slice and dice it.

I’ll be working on the premise that good forgery is very difficult to do consistently – particularly with large amounts of data. I have no way of telling whether some interns have a full time job fabricating data records to make it look real – we can only look at the data presented to us and see if any anomalies exist.
I need to advise that some earlier studies have been done on the same. I currently am aware of at least two:

1. The excellent weekly analysis of the Silver ETF bar numbers, here. We are very appreciative of the folk at http://About.Ag/ as they shared a bunch of data with us. You should check out their site for a very detailed and comprehensive analysis of all things related to silver.
2. The Project Mayhem July 2009 investigation, who tried very hard to conjure a duplicates bogeyman, but has since been really well dissected by Bron Suchecki here.

Given that some analysis already exists out there, my strategy is slightly different and there are 4 phases to this project:

1. (DONE) Construct a database and make it publicly accessible (using SQL Azure).
   
2. Make a general call out there for all historical silver bar serial number files that people might have – both from SLV and other funds.
   
3. Some kind of website component that people can put on their website to do an easy bar serial lookup (e.g. an investor could verify that a bar they received from SLV was actually held in their stock).
   
4. Over time, with enough data it might be possible to cross-reference data from various funds and check for any multiple claims on the same metal.

The most important element of this is that it’s all independently verifiable. What is there is merely a copy of what was previously publicly available. If you have TSQL skills, or you can hire a data monkey who does then you can easily check and double check the claims that I make about the claims. There’s nothing to hide here, you can test everything yourself. If you don’t have access to your own data monkey then I can help (in exchange for some benny bucks).

  Database Server: erwzbgqjg0.database.windows.net
  Database Name: BullionBars
  User Name: slvreadonly
  Password: *12345screwtapefiles



If you have access to some OLD serial bar lists which you may have downloaded in the past and would like these to be added to the analysis data, please get in contact me at this blog, we would love to have a larger data picture.

If you have some ideas for some charts and graphs, requests changes and ideas: Submit Ideas Here

Disclaimer: I don’t accept any responsibility for the accuracy of the data however it is easily verifiable from visiting the source of the original documents (listed in the database). I may change the database schema and data at any time and I may revoke access to the public login if I desire. Making the database public is NOT an acceptance that the SLV bars are real. Do not make any trades based on the information I present, however please feel free to use the information as part of your own due diligence.

Currently I have loaded data from mid-March 2011 to present, with more historical data to come (I am in possession of files from July 2010 onwards). I’ve done a little bit of analysis so far. Here is a quick profile of bar weight distribution. Dead boring – there is nothing to see here except that the shape is pretty close to what you would expect and is the classic distribution of bar size.



Then something more interesting, here is a chart of inventory by ‘country of origin’ for the refiner who produced the bars. This graph goes not currently show movement in-and-out (or movement over time), but one thing is clear – it has lots of silver from Russia and China. Again this only proves that SLV claims to have lots of silver which originates from refiners in Russia and China. Someone might be able to call the significance on that – I don’t have enough knowledge to make a judgement. Note these don’t necessary represent new bars being made, just a representation of what SLV claims to hold @ April 27.



Wading through all the data has been quite an exercise – there are a lot of records. On the surface it looks quite real to me, and the idea about JPM being long physical silver looks real when you stop to think about it. Personally we have to come up with a better theory … for example let’s say they do have the silver but can’t release it because it forms the collateral for all their leveraged paper trading then that also makes sense of what we’re seeing, and the real value for that underlying silver is worth more to them, than an 80% premium, i.e. perhaps everyone is right but for the wrong reasons. The key to this whole thing appears to be ‘derivatives’ and I’m trying to find the right interpretation of it. Discussion is encouraged and I’ll be adding extra dimensions to my analysis and database over time, but I would really welcome some notes from Amber/Wynter_Benton if you’re out there. And Kid Dynamite, just so you know – my database does not PROVE the silver is there – it simply makes the existing CLAIMS more accessible for analysis (to everyone). Paul D. Bain, I’m interested in some of your theories too if you had a comment (you owe me one).