A few weeks ago, Daniel Yanofsky published a piece in Quartz on South Africa’s import of $1,1 billion worth of gold from the USA. According to the piece, this was the reason our Q1 balance of trade had gone from hero to subzero in the blink of an eye. Like everyone else I shared this article with (in advance of the SA business press picking it up), I found the story intriguing – especially in the light of the recent gyrations in the gold market, of the stories of businessmen being stopped at the Italian border with gold ingots stuffed into their car doors, of the stories of the Chinese amassing a vast, undeclared stockpile. Not least because of the stories of price manipulation and rehypothecation. The Quartz piece seemed to pique the interest of the gold bugs too, and was almost immediately reposted by Kitco, Max Keiser… et al. Eventually the South African press caught on, but the angle was not the expected one: Moneyweb’s analysis amounted to soft-pedalling at best, and shoddy journalism at worst. Their attempts to verify Yanofsky’s JFK data had been unsuccessful, apparently. Strange… Yanofsky replied to my tweet within hours, and it took me barely 15 minutes to check his facts after I’d been pointed in the right direction. Also, Moneyweb’s “mining analyst” was a full 400 tons off in his estimation of Rand Refinery’s capacity, and a full 220 tons off in reporting the refinery’s current production levels, meaning that Rand Refinery’s spare capacity would be 800 tons. According to Rand Refinery, they have only about 200 tons spare capacity. So much for “mining analyst”. The line which subsequent reporting was to take had been established, though: “It’s business as usual, and those people asking questions are doing so because they don’t understand how things are done. Nothing to see here. Move along, please.” It was to become increasingly clear over the next few weeks that the people reporting the story knew very little indeed, and that those who were in a position to provide answers simply were not doing so.
This piece was followed in short succession by a press release from Rand Refinery, which, though it did not explicitly claim receipt of the gold, and which seemed more concerned with the ethical credentials of its product (“Our product is not sourced from any conflict afflicted areas…”), was turned into “the answer” to the riddle of the gold imports. It was from this press release that I learned something which Moneyweb’s “mining analyst” obviously did not know: that South Africa imports about 220 tons annually from the rest of Africa. None of which is used to fund conflict, of course.
But 220 tons? If our own production was only about 195 tons, then we were a net gold importer, and had been for years! Wasn’t this murdering our balance of trade?
Attempting to fact-check all of this proved frustrating, though, as it soon became apparent that even official information regarding the import and export of gold to and from South Africa is sketchy, missing, or contradictory by turns. One would think, for example, that the Department of Trade and Industry would be a good place to start when talking about imports and exports. Indeed, they have figures for monthly, quarterly and yearly imports and exports, which you can break down according to commodity. According to these figures, SA imported $29,000 worth of gold in 2006. It took just five years for the dollar-value of gold imported from the USA to reach $500 million in a year, and to become South Africa’s largest import from the USA. But when I tried to check the numbers for gold imported from the rest of Africa, I was shocked: the entire commodity line item (H7108 – Gold) was missing for the 2012 yearly figures. There were no figures whatsoever for imports or exports in that year, and the figures for previous years contained more blank fields than filled ones. Figures for Q1 2013 are probably in ZAR, but might be in USD. It would be nice if the DTI would clear this up. The figures for gold imported from the USA in 2012 and 2013 made no sense when compared to 2010 and 2011. So, obviously, the Department of Trade and Industry was not going to be the answer. Eventually, looking at the 2012-2013 monthly figures, it was possible to determine that the 220 tons of African gold processed at Rand Refinery had not found their way into the Department of Trade and Industry’s statistics.
OK, ja well no fine, maar nou wat nou? How had the American gold ended up in their export figures for 2013, but not in our import figures? By now, sources in SA were indicating that the South African Reserve Bank was responsible for keeping track of our gold imports and exports, but a trawl through the SARB website was fruitless. If the SARB was keeping tabs, they weren’t publishing the information. Meanwhile, reporting in SA was taking on an exasperated tone, though it was not, as some might expect, exasperation at how confused or obscure the official figures were. Rather, reckless “conspiracy theories” were the source of concern. A final attempt to put the questions to rest was made by Johann Barnard, writing for the “Investor’s Monthly” insert in Business Day. At least it seemed that Johann has been able to receive some kind of confirmation from the SARB, and that the reason the US shipments had not shown up in the DTI figures might be because these gold imports are classed as “temporary” so as not to distort the balance of trade. So far so good, but the last paragraph raised more questions.
While trying to get to the bottom of the riddle, reports of the COMEX gold vault’s plummeting stocks had been in the back of my mind. Could the US stuff be COMEX gold? Barnard’s article, citing the South African Reserve bank as the source of the information, seemed to indicate that the 2012 shipment was just that:
“COMEX bars (the US standard) to be refined and/or manufactured into kilo bars for the US market as per the client’s instructions.”
Once again, so far so good. Except that Howard Craig, Rand Refinery’s MD, is quoted in the same piece as saying
“All gold received is treated according to our customers’ needs and arrives at the refinery in the form of unrefined or secondary gold, and leaves the company in the form of various valueadded beneficiated gold products, such as cast bars, minted bars, minted coins and coin blanks.”
In other words, not COMEX-approved bars, which are all supposed to be 995+ purity (ie. refined gold). So, more questions then, although I doubt anyone will want to comment after Barnard’s piece, which purports to set the entire question to rest. This is especially frustrating, as Barnard’s final explanation for the discrepancy in trade figures, given by Byron Woods of ABSA NewGold, an ETF which revaulted its bullion from Rand Refinery to Brink’s in London in 2009, is that the “mystery” is the product of a clerical error on the US side. The Americans had recorded a normal “foreign trade deal”, when we recorded a “temporary import”. Which still doesn’t explain how over $1bn worth of what one assumes, if Woods is correct, would be temporary gold imports ended up as “imports” in the DTI’s figures for 2010 and 2011, while the latest figures don’t reflect anything at all.
Perhaps it’s all too much for the DTI. Or perhaps they didn’t get the memo. Finally, I came across this little gem, in the minutes of a 2010 NCOP committee meeting, and reported to be the utterances of the then Acting Chief Director of the Department of Trade and Industry:
“South African gold and diamonds were sold through third party markets and were excluded from export statistics…”
And then, tellingly,
“Gold was exported to Frankfurt, Germany and all countries that received gold from South Africa obtained it through Frankfurt…”
And so, a possible (though increasingly surreal) scenario began to emerge, explained quite succinctly by Riaan de Lange at the University of Pretoria: before June 2008, gold was imported into bond, and, it seems, not declared as an import. In July 2008, it was decided to change this, and gold imports were now included in the calculation. Gold exports, however, were still excluded. Then SARS decided not to include gold imports, after the calculation of the trade balance was adversely affected, since the exports had not been included. Still. Ergo: No imports, imports, no imports. In the space of a few months.
Now we’re supposed to believe that something similar has happened all over again, less than 5 years later.
So what have we learned?
1. At least we know where (some of) the COMEX gold is. Or not. For now. Depending on who you believe.
2. The amount of gold entering SA from the USA is skyrocketing. If this continues, the likely next move will be for the US to stop declaring these exports, in the same way that we don’t declare ours (unless this was just someone ticking the wrong box, in which case we could have been importing large amounts for years without anyone noticing). Because, obviously, if your product is destined for a “third party”, you don’t have to declare it.
3. We don’t declare imported gold in our import figures, but when SARS compiles the balance of trade figures, it is included, as a separate item. Good luck trying to find out where it is all coming from…
4. Apparently, we don’t declare it as an export, either. So good luck trying to figure out where the stuff goes once it leaves Rand Refinery. (Hint: “Third Party”, Frankfurt)
5. And all this non-declaration is IMF SOP. So STFU.