How can we trust the Bar List Data?

Recently, video blogger Alexiscom1 did an analysis of some of our material and asked some questions link. Questions are good - they help us interpret the world around us. One of his questions I thought would make a good post: (paraphrased) "How can we trust the Bar List Data?" Alexiscom1 says: "... any report from SLV in it self would not worth much if they are doing their own audit them self." A popular sentiment! Also talked about by Dave in Denver: 'Any reports generated by SLV/GLD are only as good as the paper they are recorded on.' link. The bar lists are my data source, so I better talk about it at least once.



First, just to set the scene - short of strapping a camera to a kindergarten child's head going to an organized SLV vault day tour, we are limited in our ability to get direct first-hand proof of the inventory. I have always loved this printer ad of the martians printing out boring mars pictures to put in front of the camera of the mars rover. The crux of the ad is that the quality of the print is such good quality that the viewers mistake it for reality.



The ETF's each independently (publicly) publish information about their inventory and present it as evidence of their holdings. I am simply collecting that data and storing it in a central archive for study and analysis - nothing more complex than that. So the bar lists are the published set of claims from the ETF's, how can we trust them?

To address this adequately, we need to recognize it is actually several questions in one, only then can we start to construct tests. Interestingly, the questions are progressive, which means as soon as we answer the first, we can move onto the next 'level' which is typically more complex (conversely it is difficult to move down the ladder). Identifying the categorization may help discussion, so here's the strata:

Level 1: Does the bar list represent bars in the vault? (i.e. empty or not?) Turd Ferguson says largely empty, Harvey Organ says completely empty.

Level 2: Does the bar list represent bullion which is not fake? Bix Weir suggests it's most likely all Molybdenum or Tungsten, Rob Kirby alleges salted gold bars are Everywhere. These people generally call for 'worldwide re-melt and assay'. Perhaps they are correct! But we do at least have a way we can data hunt for these fake bars if they do exist (another post for another time).

Level 3: Does the bar list represent real bullion which is not being double-counted? Ah, the age old con - selling the same thing more than once. This isn't a very common argument but it does pop up continually like with folk like David Morgan, or maybe even GATA. For the purposes of this definition relating to ETF holdings, 'Double-Counting' is defined as ownership - another entity AND the ETF both having full title to the same bar at a serial number level (regardless of how it came about). After much email discussion it turns out that this kind of situation would be rare - inasmuch as the wording on contracts regarding metal would not typically be written to include specific serial numbers, and that bullion banks would not be so stupid to perform that trick on bars which they know are in the public record (in the bar list). For this situation though, we do have a practical test! If you do hold full title to bars by serial numbers, we have the ability to do an ETF spot check for you - on the serial number, using our database. Your evidence combined with ours, might crack this case wide open!! Sadly, no one has yet stepped forward. This absence of people with numbers to test does not prove that there aren't any double-counted bars in ETF's but it's not for lack of opportunity!!

Anything past this level starts to get complicated, and cannot be addressed by the bar lists by themselves. I will however, attempt to itemize them. Please note that a higher-level question only represents argument complexity.

Level 4: Does the bar list represent real bullion in the vault which is not leased? Spending some time agonizing over this definition, I am taking the easy way out and just leave the definition of 'leased' as a broad classification, distinct from level 3 argument. There are of course, many conversations surrounding leasing gold or silver and I will leave the discussion for others. Some reliable starting points for this discussion is Bron Suchecki's 2010 article about GLD, leasing and encumbrances, and Kid Dynamite's 2011 Today In Silver Misinformation: David Morgan Swap/Lease Edition which is a cracker - especially the comments section. Discussions at this level will invariably see Hypothecation or Dave in Denver being flung around. We could put GATA in level 4, along with any argument which suggests central bank stockpiles have been leased onto the Bullion Banking system. I don't have any solutions for proving or disproving discussion at level 4, but in theory if we had good data on other bar stockpiles (from outside of the ETF system) then we might be able to do some data matching and put together a picture. My contribution (for other investigators to address) is a big stash of historical bar list records.

Level 5: Does the bar list represent real bullion with legal claims, but connected with additional hidden fraud? The thrust here is "yes, the inventory is real with unencumbered metal, fairly represented by the lists, but the fraud is elsewhere". The definition of financial fraud may vary within this definition and should not be conflated with 'risk'. Anyone arguing in this space will have agreed that everything's fine in relation to inventory, claims and leasing, but that some kind of connected fraud is still taking place such as price manipulation OR High Frequency Trading or social media obfuscation.

Level 6: Does the bar list represent a legitimate financial instrument (ETF/Trust)?
Kid Dynamite operates at this level (and says yes), so does Jeanne D'Arc. This is the highest complexity argument because it assumes that all the arguments put forward in the previous levels have been addressed in some fashion, typically through experience, industry knowledge or research. It is also worth noting that many people entering metals for the first time will probably start here but if they are unable to mentally negotiate their exposure to internet metal-memes they may get dragged down to the less-complex arguments.

 

Typically if we prove levels 1 and 2, then the argument moves automatically to level 3 and so on, so you can see that everyone looking at the silver/gold market is on their own mini-journey. The effect is subtle, just like someone who writes with incorrect spelling or grammar will be largely ignorant of their mistake unless they critically analyze their own opinion with spellchecker - whereas it is instantly noticeable by someone with perfect grammar.

Breaking it down like this can tell us a lot about the level of knowledge and research done by each individual as well as their exposure to internet discussion. Alexis' question indicates that he is somewhere in level 1 and 2 arguments. For social media obfuscation, a common trick seems to be loading several levels of argument all into one bleat, which understandably has the effect of confusing or overloading the reader. While it is possible that many types of fraud take place simultaneously (i.e. manipulating the price on a leased, fake, non-existent bar), I can think of many reasons why combinations would break down at some point.

Anyway, good news Alexis; currently we can reliably vouch for at least 1% of SLV's bars being real - it doesn't seem much but we've got plans for doing even more investigations so stay tuned!

The concise answer to Alexis' inquiry is 'we cannot trust the bar lists implicitly, but we can test their consistency internally (within the data) and externally (against what we know of the bullion industry) and address questions at level 1, 2 and 3. Here at Screwtape Files we are simply designing tests we can use to look at the data. At some point, everyone's quest for evidence must simply give way to practical things. It is not not practical to test the purity of every single bar in existence - in exactly the same way you would not bother to ensure that the pallets of cola in the back room of the bottle-shop were real. Bad analogy ... I mean to say that everyone carries a set of specific proofs necessary to satisfy inquiry, and only when those thresholds are met can we move on. These proof requirements are always different for different people and will also change over time, experience and circumstance. For myself I've spent countless hours putting together investigations like SLV Database 4, and sort of in vain because all I had to do was read slide 10 this LBMA presentation (about bar manufacture quality) link where it states quite clearly, 'the bars move between vaults'. Nowadays I can vouch for the claim backed by data, whereas normally this kind of stuff will be ignored by the standard conspiracy arguments because it makes the 'fake bars' level 2 argument much more difficult to uphold.

My critic David suggests I am wasting my time with the bar lists. I actually agree - as you can see above, it doesn't allow us to investigate levels 4, 5, and 6 (where fraud becomes lucrative). This is purportedly his argument level so I was completely surprised to see him drop a 'fake bars' level 2 argument in his last comment. But since other people are starting at these levels I hope to supply them with well-structured stuff to challenge their views. Anyway, that's my small offering for today - a framework for interrogating people's approach to bar list integrity. Please feel free to add to the definition in the comments below, we'll see if we can refine it.