TreasureHunter said:
The 9,000 tonnes and close figures appeared in a large number of blogs and writings, I read that on CNBC before I posted this question, but I also found it on SeekingAlpha and a number of other sources that are often quoted as "credible" in the PM community.
So I'm convinced now that there are lots of myths and who knows what else is inaccurate and who know what else we believed in the past simply isn't true.
Official sources and the WGC are most credible.
Not that large number, as the web search results showed: 3950 versus 196.
There might be a conclusion to draw here: that the places you visit, started to have their own version of reality. One may wonder about the level of criticism their readers show. Just accepting without verifying any of it. See, you started this topic by already wondering where those 800 'missing' tonnes went to. So it's like you said: "I think pretty much everyone is quoting everyone without much verification. ". I ceased to read Zerohedge etc stories altogether. It's just a waste of time. And I only know due to experience, having seen that they give bogus / incomplete stories.
The same bogus stories, repeated over and over again, with apparently nobody discovering that these are bogus, year after year. I'm active in this 'pm world' since 2011. That's just 2 years ago. There are plenty others out there that are much longer active in this 'pm world'. Yet, it's like they never checked anything of what they have read. The best example is the story that compares daily trading volume with mine production alike one can only buy/sell 1 time in a trading day. Repeated/copypasted all the time. You surely have read it.
The last one I discovered as bogus, was the one about the silver market being a tiny small one. I had read that so many times on Kitco and elsewhere, that it just seemed to be a universal consensus. But on a certain day, some months ago, I decided to actually check it. Well, it was true before 2002. Since, it became 160 times bigger.
As also said, I was surprised to see Seekingalpha publishing the wrong figure.
One may wonder where the higher number origined for its first time.
And to add to the complexity, as you suggested, data is subject to revision. Revision is also normal. For ex, the central banks monetary data is almost always updated a month later / final figures. But the update is in the order of fractions of percents, and only the latest figure rivised.
But apparently, some crucial research companies, in this case that UK company that provides the WGC and the SI production demand/supply worldwide data, revise numbers many years later.
How I know? Because I downloaded the data last year, I downloaded it now, and I see that figures as far back as 2002 are changed with sometimes 20%. And not a few, but dozens out of hundred. That is absurd and troublesome. I made a topic somewhere on this forum about it. And why: because I can't even guess a reasonable explanation for a whole part of the data (if not all) being revised a decade later. Imagine that I had been active in the PM world since 2000. In 2002 I save some data from the SilverInstitutes website, and I use it in posts on the forum. 10 years later, the data is adjusted 20%, and my figure had been wrong for a whole decade. While I just saved it from their site. That's just... ridiculous. Because why on Earth would someone think he needs to check every month if the figures of a decade ago weren't revised / are still the same? In the case of the SI data, I only found out by coincidence, I was checking the data for my post about the size of silver market increase, since the SI only shows the last 10 years data, for 2001 and 2002 I had to use an in 2011 harddisk-saved copy, and while processing the data, I noticed differences everywhere. I then searched the web for older copies of SI's supply/demand, found some, and these confirmed that my harddisk-saved copy indeed wasnt edited somehow.
Aside of this, I don't give any site any creditibility. Even not official ones. Aboves SI example is an example of an official site giving wrong data for almost a decade (or, also possible, giving wrong data since the revision). Instead, I try to collect as much data I can find. And I store it as new files/lines, never replacing older, just adding. From the whole of it, I try to get an idea of how things actually are. This topic does a good job as a warning shot.
