Since 2015 I’ve stated the raw data published by the Australian Bureau of Statistics (ABS) on
gold export to China mainland do not accurately reflect what is imported
by China from the land of down under. In my previous
post I’ve written Australia’s gold “export to China” should be
adjusted by Hong Kong’s gold “import from Australia” as a substantial amount
of gold Australia declares to export to China is in fact travelling through
Hong Kong. A written statement from ABS now supports this analysis.
Obviously, what we’re after is exactly how much (non-monetary)
gold China is importing from all countries around the world. As
China doesn’t disclose its cross-border trade statistics for gold, the only
way to reach our objective is by establishing net gold export from all
countries around the world to China, subsequently to aggregate these numbers
and come to an estimate of total Chinese gold import.
One of the largest exporters of gold to China is Australia, but
calculating Australia’s gold export to China is complicated, as we need to be
cautious to avoid double counting in computing China’s total net gold import.
The thing is, ABS bluntly alters any “export to Hong
Kong” into “export to China” when they suspect the gold is shipped to Hong
Kong but will reach the mainland as its final destination – this is what ABS
confirmed to me. Therefor, from using Australia’s “export to
China” double counting can arise when any gold flowing from Australia via
Hong Kong to China would then be declared as “export to China” by Australia
and “export to China” by Hong Kong.
Let us clarify this subject. Have a look at the chart below. If we focus
on the period from January 2013 until December 2014 we can observe that
Australia’s “export to China” (purple bars, data sourced via ABS/COMTRADE)
roughly matches Hong Kong’s “import from Australia” (turquoise bars, data
sourced from the Hong Kong
Census and Statistics Department). We must conclude that over this period
roughly all gold Australia exported to China was transferred via Hong Kong.
If we then would sum up Australia’s gold “export to China” to Hong Kong’s
gold “export to China” this would result in double counting.
Exhibit 1.
Australia’s “exports to China” versus Hong Kong’s “import from Australia”.
All data by ABS I’ve sourced via COMTRADE simply because it’s cheaper.
Crosschecks between ABS and COMTRADE show their data is nearly exactly
the same.
Historically, Hong Kong has been the main entry point for gold into China.
Not surprisingly Australia’s gold “export to China” was all directed through
Hong Kong until December 2014, as we can see above. The situation changed
early 2014 when China openly stated to stimulate direct gold imports from all
over the world, bypassing Hong Kong. Soon after, we’ve witnessed the
birth of direct gold export to China from for example the UK.
Exhibit 2. The UK’s
cross-border gold trade with China mainland. Halfway 2014 the UK commenced to
export gold directly to China mainland bypassing Hong Kong. Side note: since
the beginning of 2016 the UK’s export to China has died out on a rising price
of gold, the
UK turning into a net importer of gold.
The strategic move from Beijing to stimulate direct import was also
mentioned on Chinese state television channel CCTV in 2014, see the clip
below starting from 1:30.
Exhibit 3. Courtesy CCTV.
Tellingly, what Australia declared as “export to China” started to
exceed what Hong Kong declared as “import from Australia” (exhibit 1) in
early 2015. The difference reflects Australia’s direct export to China,
which took off in January 2015. To compute Australia’s direct export to China
we should subtract Hong Kong’s “import from Australia” from what Australia
declares as “export to China”.
When I wrote ABS about my general findings displayed in exhibit 1 they
replied [emphasize en brackets by me]:
… I can confirm that we are reporting basically as you described.
The ABS developed a policy some time ago in relation to reporting
country of origin [final destination] for gold exports. The
policy dictates that some trade with a reported country of destination as
Hong Kong, should be corrected to China. This was based on
our investigations at the time, which found that although Hong Kong is a
central hub for gold trade, the goods are predominately sent elsewhere in the
Asia/Pacific region. We also confirmed that in some cases the
destination beyond Hong Kong was not known by the exporter at the time of
export. China was assessed as the most
likely destination …. We acknowledge that this blanket
approach is not ideal, and can lead to the series differing from
expectations.
We will continue to report in this fashion in the foreseeable future.
However, as some time has passed since the initial investigation we will be
conducting a review of the policy discussed above. We will be sure to inform
you of any outcomes resulting from the review process.
Exhibit 4. Quote from email of the Australian Bureau of Statistics
(ABS) to Koos Jansen.
Case closed? No, I have a few more thoughts I would like to share.
To come to the most precise amount of bullion net exported from
Australia to China mainland we should set up a formula. We have a
few flows of bullion between Australia, Hong Kong and China to work with.
Some are known and publicly disclosed, some are unknown but can be
calculated:
- We have publicly disclosed by ABS/COMTRADE what
Australia declares as net “export to China” (C).
- Effectively, what ABS declares as net “export to China”
(C) either is a direct net export to China (Y)
or exported to Hong Kong but disclosed by ABS as export to China (X)
as they guess the final destination is the mainland (so C = X +
Y). What we want to know is the partition of Y
and X, but this is unknown.
- Then we have also publicly disclosed by ABS/COMTRADE
what they consider Australia to net “export to Hong Kong” as a final
destination (B). This amount is small but relevant for
our formula. (Actually, for us it is not important what ABS thinks the
final destination is, as long as we know this gold is arriving in Hong
Kong.)
- Finally, we have publicly disclosed by the Hong Kong
Census and Statistics Department (HKCSD) Hong Kong’s net “import from
Australia” (Z).
What we want to calculate is Y (Australia’s direct net
gold “export to China”), but the only figures published are C
and B by ABS, and Z by HKCSD. Fortunately,
we can calculate Y from C, B
and Z.
We know Hong Kong’s “import from Australia” (Z) equals
Australia’s “export to Hong Kong” (B) plus Australia’s
export to Hong Kong data that is altered into “export to China” (X).
ABS may alter the data but nevertheless the physical gold flows to Hong Kong.
Therefor:
Z = B + X
As Z and B are publicly disclosed, from
here we can calculate the amount ABS alters from “export to Hong Kong” into
“export to China”, which is X.
X = Z – B
By knowing X we can finally calculate how much gold is
directly net exported from Australia to China (Y).
C = Y + X
Y = C – X
Y = C – Z – B
By using this formula Australia net exported 113.81 tonnes of gold
to China mainland in 2015. In contrast, GFMS that wrote in their Gold Survey 2016 Australia
exported 187 tonnes to China in 2015, which is the raw data disclosed by ABS
(Australia’s gross gold “export to China”). The numbers by GFMS cannot de
accurate.
Exhibit 5. Quote from
the Gold Survey 2016 by GFMS.
One final point. How do we know the gold exports from Australia to Hong
Kong are re-exported
to China mainland and not for example to Thailand?
Re-exports are products which have previously been imported into Hong
Kong and which are re-exported without having undergone in Hong Kong a
manufacturing process which has changed permanently the shape, nature, form
or utility of the product.
Exhibit 6. Definition of re-exports by HKCSD.
Well, next to the fact ABS isn’t always sure what the final destination is
of gold exported to Hong Kong, neither do I. But, once the gold has arrived
in Hong Kong we shouldn’t care about ABS data, we must rely on HKCSD’s data.
The HKCSD states Hong Kong to re-export hundreds if not thousands of tonnes
of gold per year – Hong Kong is one the largest gold trading hubs globally.
Therefor, I assume they accurately declare how much imported gold is
re-exported. If the HKCSD declares so many re-exports why wouldn’t they
accurately declare re-exports from Australia to the mainland?
Eventually, by calculating Australia’s direct net gold
“export to China” (Y) through the formula above, and adding Hong Kong’s
“export and re-export to China” from HKCSD data we should have the
best assessment of China’s net gold import from Australia and Hong Kong.
Exhibit 7. Hong Kong’s cross
border gold trade, which in my judgement is one of the most accurate data
sets globally. Side note: bullion that was accumulated in Hong Kong during
2013 and 2014 is flowing out since 2015.
My assessment for China’s total net gold import for 2015 is 1,565
tonnes. If we add domestic mine output at 450 tonnes
and a few hundred tonnes of scrap supply and disinvestment, total supply has
been well over 2,000 tonnes for the year. Not surprisingly SGE
withdrawals accounted for 2,596 tonnes in 2015.
I know, I owe you all a respons to the Gold Survey 2016 by GFMS in
which they stated Chinese gold demand in 2015 was 867 tonnes – quite a gap
with SGE withdrawals. As some of you know, or might have guessed by now, I’ve
had some health issues in recent years hence my writing hasn’t been constant.
Hopefully, things are turning for the better again and I can continue
publishing on the Chinese gold market (my ongoing disagreement with
WGC and GFMS), global gold flows and a whole bunch of other planned articles.
