What about Sprott's 4,500 tonne export 'gap'?

Someone emailed me yesterday:

FOFOA,

I was reading Turd Ferguson's blog the other day and he brought up a pretty interesting point about Western CB's not having any more gold left (based on Erik Sprott's research regarding the gold exports from the US) and it's impact on Freegold.

Would you be so kind as to share your thoughts on this topic?

I've included the link to his post below for your reference.

http://www.tfmetalsreport.com/blog/5041/fifty...dollar-question

Regards,
Paul

I read the article and followed the comments below it, Motley Fool's in particular. I see that he is very popular over there! ;D I also got a kick out of this comment by Nickelsaver:


One of the primary hypotheses of the gold bug/GATA camp that stands in stark contrast to A/FOA says that the Western central banks have surreptitiously emptied their vaults supplying physical gold to the market. The TFMetals article highlights this contrast in particular, pointing to the latest "smoking gun" which has been circulating for about six months since Eric Sprott wrote target="_blank" Do Western Central Banks Have Any Gold Left??? Part II last March.

The "smoking gun" is the net U.S. gold exports spanning 1991 through 2012 as reported by the U.S. Census Bureau which Eric Sprott says exceeded what the U.S. should have been capable of exporting based on supply estimates which excluded private sales as "unknown." The amount of gold exported in excess of what should have been possible (not counting private sales) was 4,500 tonnes over the 22-year period:

"We used this framework to analyze supply and demand in the US going all the way back to 1991, which is as far back as the FT900 documents go. Over the span of 22 years, the total amount of gold that the US has exported – above and beyond its supply capability – is almost 4,500 tonnes! A truly stunning figure. (See Table 3)."

Sprott discounts the possibility that the extra 4,500 tonnes could have come from private sales by claiming that Western gold investors have been net buyers over the last 22 years:

"Admittedly there is an unknown in our analysis, that being gold bullion acquisition and disposition by private investors. However, strong demand in ETPs such as GLD and PHYS and demand for gold coins provide strong evidence that the private investor has been a net buyer over the years. The inclusion of the private investor on the demand side would in fact skew the ‘gap’ of 4,500 tonnes higher to a figure that would lie somewhere between 4,500 tonnes and 11,200 tonnes, which represents the gross exports out of the US. The only US seller that would be capable of supplying such an astonishing amount of gold is the US Government, with a reported gold holding of 8,300 tonnes. The US Government gold holdings have not been audited or verified in more than four decades. The US trade data defines the export of nonmonetary gold as a sale of gold from a private seller within the US to an official agency. In September 2012, we espoused that the Western Central Banks have been surreptitiously selling/ leasing their gold through private channels in an effort to increase the available supply and in turn suppress prices. This new analysis using official US agency numbers seems to provide the strongest validation of our hypothesis to date. It is worth noting that our data only covers two decades and that the export ‘gap’ could in fact be significantly larger if earlier numbers were included or the real private investor demand for gold was known."

Note that this contradicts what Another said about Western gold investors trading their physical for paper gold.

Motley Fool did a good job addressing this weakness in Sprott's analysis in the comments over there:

Submitted by Motley Fool on September 10, 2013 - 7:36am.

I went ahead and read the only 'proof' you offered, being the two part paper by Sprott et al.

Sufficed to say the analysis is very weak, and has huge gaps.

Their figure for the 2012 disparity was 50 tonnes. The average for the period was just over 200 tonnes. The timeframe as 1991-2012.

They do not link the exact data, but this implies that most of the disparity came from the earlier period, rather than the later one.

Now think about gold and sentiment regarding it in the general public sphere from say 1991-2000. Horrid would be a good word. Hell, even today, despite a bull run the general public sentiment is an abhorrence for the barbarous relic.

Their weakest point in analysis is private supply/demand. I think it very feasible that that amount could have been supplied from private hoards, over the period, especially the first part ( where I am guessing most of the discrepancy arises).

Furthermore no differentiation is made between physical bullion, and say unallocated spot. On the demand side many 'sophisticated' investors over this period would have opted for the latter, which is in effect no demand at all. This lack of differentiation also skews the analysis.

Did the US public have the wealth to supply this differential? Yes.

I think this is the more likely explanation for the remainder in difference, once one makes a distinction between paper gold and real gold.

The author of the article, "Pining 4 the Fjords", then defended Sprott's claim that 4,500 tonnes is too much to have come from the American public alone:

Submitted by Pining 4 the Fjords on September 10, 2013 - 8:49am.

…Third, you correctly mention the possibility that this could have come from private sources. Indeed, Sprott et al noted that this was a possibility in their second article. However, they are positing that this much gold is too large to have come from private sources. So the question is, how likely is it that 4,500 tons would have come from private owners of gold in the US? Here is the list of gold reserves

Rank Country/Organization Gold (tonnes)

1 United States 8,133.5

2 Germany 3,390.6

3 IMF 2,814.0.

4 Italy 2,451.8

5 France 2,435.4

6 China 1,054.1

7 Switzerland 1,040.1

8 Russia 1,002.8

9 Japan 765.2

10 Netherlands 612.5

Now here is a list of the top ten largest holdings of physical gold in private hands worldwide:

Privately held gold Rank Name Type Gold (Tonnes)

1 SPDR Gold Shares ETF 1,239

2 ETF Securities Gold Funds ETF 259.79

3 ZKB Physical Gold ETF 195.53

4 COMEX Gold Trust ETF 137.61

5 Julius Baer Gold Fund ETF 93.50

6 Central Fund of Canada CEF 52.71[14]

7 NewGold ETF ETF 47.75

8 Sprott Physical Gold CEF 32.27

9 ETFS Physical Swiss ETF 27.97

10 Bullionvault Bailment 37.1[15]

Please note that 4,500 tons would be more than 2x larger than all of the 10 largest worldwide private holdings of gold combined! How likely is it that from 1991-2012 private owners of gold in the US sold the equivalent of what would be the second largest gold hoard in the entire world? Is it likely that they would sell (or even own) more than the top ten worldwide privately held gold hoards combined? I think not. That is why Sprott was saying they only reasonable source for this much gold was the US treasury. I still believe this is a reasonable conclusion…

MF then countered:

Submitted by Motley Fool on September 10, 2013 - 11:49am.

As to the first, yes I appreciate it.

As to the second and third.

No. This is their assumption.

"However, strong demand in ETPs such as GLD and PHYS and demand for gold coins provide strong evidence that the private investor has been a net buyer over the years."

I call BS.

"Please note that 4,500 tons would be more than 2x larger than all of the 10 largest worldwide private holdings of gold combined! How likely is it that from 1991-2012 private owners of gold in the US sold the equivalent of what would be the second largest gold hoard in the entire world?"

Yes. Over a period of 22 years. I think it is fairly likely. If he would, I would appreciate some input from 'Nick Elway' who I know comments on these forums as to the possibility US citizens held and were able to sell that much gold privately. He has some of the best data in the business, and I think his input would be valuable.

In their study they simply say that private supply/demand are unknowns, and proceed to ignore them. Since they provide no breakdown of the import/export data, it isn't possible to check, but are we then to deduce that they conclude that over the period from 1991-2012, no private holder of gold sold any of it to someone else outside US borders? Seriously? Not even an ounce? All of it 'must have' come from CB stockpiles?

And you wonder why I call their analysis flawed and weak. :P

…Much of this lies in the realm of speculation, as very few hard facts are available. Unfortunately we may not known the truth, till the chips are down. That said, the 'western banks have wasted all their gold' idea does make some assumptions which I consider insane, such as central banker stupidity.

MF

I documented this discussion here because I want to present another possible explanation for this 4,500 tonne "export gap," or at least part of it, that I haven't seen anyone acknowledge in the six months that Sprott's analysis has been circulating. As it is framed by Eric Sprott and Shree Kargutkar, his co-author, there are only two possible explanations (or a combination of the two). It was either public sales (unlikely says Sprott) or it was physical transfers (sales or leases) from the central banks into the physical market, with the implication that at least some of it would have come from the U.S. stockpile since we're talking about U.S. exports reported by the U.S. Census Bureau.

The Federal Reserve Bank of New York (FRBNY) holds mostly foreign gold and only a small portion (about 5% or 418 tonnes) of the U.S. stockpile. The rest of the U.S. gold (7,715 tonnes) is at Fort Knox, West Point and the Denver Mint. The FRBNY also reports how much foreign gold it is holding as " target="_blank";Earmarked gold". In his post on target="_blank" Central Bank Gold Leasing last November, Victor The Cleaner made a chart of foreign official gold held at the FRBNY from 1982 through 2012. This is Exhibit 1 in my alternative explanation:


As you can see from the chart, close to 4,000 tonnes of foreign official gold left the FRBNY between 1991 and 2012. Could that gold have shown up as exports in the Census Bureau data? Exhibit 2 is a conversation found in the minutes of a Federal Open Market Committee (FOMC) meeting on December 22, 1992. The Fed releases these minutes 5 years after the meeting:

CHAIRMAN GREENSPAN. Did I hear you correctly when you said that the gold exports in October appear to have come from the coffers of the Federal Reserve Bank of New York? Has anyone looked lately?

MR. TRUMAN. Well, I didn't want to tell too many secrets in this temple!

VICE CHAIRMAN CORRIGAN. Obviously, we knew what happened to the gold, but I don't think we knew what it did to exports.

MR. TRUMAN. What happens in the Census data is that the Federal Reserve Bank of New York is treated as a foreign country. [Laughter] And when a real foreign country takes some of the gold out of New York and ships it abroad, it counts first as imports and then as exports. However, the import side is not picked up in the Census data. So there you get the export side of it.

MR. LAWARE. Great accounting!

MR. BOEHNE. Great confidence building!

MR. TRUMAN. That's because you haven't been filling out your import documents!

MR. ANGELL. Let me run this by again. You mean a country owns gold and has it stored in the Federal Reserve Bank of New York and if they ship it out, that's an export?

MR. TRUMAN. And in the balance of payments accounts it also counts as an import, so it washes out.

CHAIRMAN GREENSPAN. The Federal Reserve Bank's basement is a foreign country. When they move it out of the basement into the United States, it's an import. Then, when they ship it out again, it's an export.

MR. ANGELL. That makes sense!

MR. TRUMAN. And sometimes when they sell the gold, it might be sold into the United States, so it should count as an import. It doesn't necessarily always show up as an export.

MR. BOEHNE. That really clarifies it!

MR. KELLEY. Does it have to get out of your vault at all in order to be considered an import and an export?

VICE CHAIRMAN CORRIGAN. Well, I'm not even going to try to answer that. In this particular case I know what happened, so I think the description you have is correct.

Source: target="_blank" http://www.federalreserve.gov/monetarypoli...1222meeting.pdf

[Credit for dredging this out of the old FOMC minutes goes to Adrian Douglas of GATA who passed away this year. gata.org/node/8429 target="_blank"gata.org/node/12137]

I want you to notice a couple of very interesting things in that exchange. First of all, Truman says that in the BOP accounts it counts as both an import and an export so it "washes out," but in the Census data it shows up as a net export. And then he also says that if it is sold into the market in New York, like say to the bullion banks like JPMorgan, it would actually show up as an import. Extending that logically, if it was CB gold coming out of the FRBNY being sold into the market and then exported (which is Eric Sprott's theory), it would show up as an import first and then an export and it should be a wash rather than a net export. So the only way it should be a net export is if the foreign CB is taking it home and there was no transfer of ownership or sale into the market.

And finally, in the last two lines, it almost sounds like there may be cases where gold could show up in Census data as an import or export without ever leaving the FRBNY. So there's certainly plenty of room to question the Census data, especially as it relates to central bank gold. Yet Eric Sprott doesn't seem to have even considered this issue in his articles. Was he not aware of these FOMC minutes published by GATA three years earlier?

The approximately 4,000 tonnes of gold that left the FRBNY during those 22 years was not American gold because it came out of the "earmarked gold" which is simply custodial gold held on behalf of foreign governments and central banks. So there's no reason its movement should show up as "monetary gold" on any U.S. national account for any purpose. The only thing that makes sense in light of the FOMC minutes above is that physical transfers of foreign gold (as opposed to transfers of ownership) are being treated as nonmonetary gold movements for the purpose of import and export reporting. And if that gold was sold into the market, then it should have shown up as either a net import or a wash rather than a net export. But I can easily imagine how a CB gold repatriation (which is most likely what the 4,000 tonnes leaving the FRBNY was), especially if it was done in secret, might show up as a nonmonetary gold export in the Census data.

Thanks to releases by the Bundesbank over the past year, we now know that Germany moved 930 tonnes from London to Frankfurt in secret and only revealed the move ten years later. From a Q&A posted on target="_blank"the Bundesbank website on October 25, 2012:

"At the beginning of the last decade, we brought 930 tonnes of gold to Frankfurt from London and subjected it to a painstaking inspection. Part of the gold was melted down in order to create new bars which conform with the “Good Delivery Standard” which is customary nowadays in gold trading. Of the 930 tonnes of gold, not one gram was missing. We do not have the slightest doubt that our holdings in New York and Paris are also made up of the purest fine gold. We have at our disposal fully documented lists of the bars, and our partner central banks send us every year confirmation not only of the bars’ existence but also of their quality. We receive confirmation of our gold reserves, measured in troy ounces. The Bundesbank has been drawing up its accounts on this basis since it came into existence. All external auditors have confirmed our accounting practices outright since then."

Moving hundreds of tonnes overseas is serious business. You can imagine why it would normally be conducted in secret. Whenever physical gold is in transit, it is subject to loss by theft or even accident. Remember target="_blank"the story about the Indian van that broke down on the freeway while transporting official gold to the airport? And this isn't the old Bretton Woods era where national treasuries and central banks were the most adept at moving gold. Today is the era of the LBMA, when bullion banks like JP Morgan and private transports like Brinks and VIA MAT are the experts at transporting gold bullion.

So I can imagine that a CB who was repatriating some of its gold from the FRBNY might contract with these private firms to have its gold fully insured through the private sector while in transit, rather than pulling an Indian van stunt. And this would mean the technical "demonetization" of the gold during transit such that it would show up as nonmonetary gold the moment the FRBNY handed it over to Brinks or JP Morgan. And as Edwin M. Truman said above, that would technically be a gold "import" as it left the FRBNY but it wouldn't show up on the Census data; it would only show up as an "export" when it left the country.

It's actually quite interesting to think that the Fed treats these CB repatriation transfers as nonmonetary gold movements. It certainly makes sense from the $IMFS perspective! And it also tells me that they are likely transporting via the private sector, like through JP Morgan who target="_blank"has direct tunnel access to the FRBNY. That would be the "import" when it is moved off of Fed property into the tunnel, and then when it leaves the U.S. via JFK airport would be the "export" which should theoretically "wash out," but as Truman, a Fed economist, told the Fed bureaucrats in 1992, it shows up only as a net export "because you haven't been filling out your import documents!"

So the Fed bureaucrats don't bother with the "import" documents but JP Morgan or Brinks or whoever is flying it to Europe does fill out the export documents and it shows up in the Census data rather than washing out!

MR. ANGELL. That makes sense!

MR. BOEHNE. That really clarifies it!

MR. LAWARE. Great accounting!

Eric Sprott's research found that gross gold exports from 1991 through 2012 were 11,223 tonnes. Some of that must have been movements out of the FRBNY, else why would Edwin Truman have been explaining to Alan Greenspan how gold exports in October of 1992, appearing as part of the U.S. balance of trade, appeared to have come from the coffers of the FRBNY?

We certainly have a mystery here, but we also most-definitely have a potentially significant source of gold "exports" that was not even considered in Sprott's analysis. And by potentially significant I mean potentially accounting for up to 88% of Sprott's "export gap" which means, given an expanded statistical margin of error, potentially accounting for all of it.

By expanded statistical margin of error, I'm referring to two things. The first is that Sprott's analysis of the Census data had to convert currency terms into weight terms. The Census Bureau reports exports monthly in currency terms, so his analysis had to assume a monthly average for the volatile price of gold in its conversion delivering a considerable margin of error. The second is th target="_blank"at, apparently, the Census Bureau receives some lower purity nonmonetary gold export data from the exporters in weight terms and then bureaucrats systematically convert it into questionable currency terms, and then when the gold bugs convert it back to weight it could be off by an even more considerable margin.

Yet while Eric Sprott's 4,500 has a considerable margin of error, the ~4,000 tonnes that left the FRBNY does not. It is also reported in currency terms, but no averaging is needed because the price of gold repor target="_blank"ted here was frozen at $42.22 per ounce for the entire period!

< target="_blank"div>