PM Daily Market Commentary - 10/7/2013

davefairtex
By davefairtex on Mon, Oct 7, 2013 - 11:26pm

Gold closed up +11.20 to 1322.40 on light volume, with silver up +0.60 to 22.35 on moderate volume.  The gold/silver ratio dropped a big -1.13 to 59.17.  Silver was clearly leading today; after trading sideways in asia and mildly higher in europe, silver broke through resistance at 0942 EST on a massive 28 cent 2500 contract volume spike.  The one-minute spike took silver through the top of its trading range, sending shorts running to cover.  Gold followed, but its move was substantially more tepid, moving only $4 during that same timeframe.  I could not associate the move in silver with any news items, and no other instruments moved up or down at the same time, so I'm writing this one up to "upside manipulation" - i.e. stop-running by the big guys.  Nobody buys 2500 contracts and moves the markets this much if their objective is to buy at the lowest price possible.  Silver kept much of its gains by the close, but gold did not, with silver closing the day clearly above its 50 day MA, which is all quite bullish for silver.

The buck dropped today, down -0.28 [-0.35%] to 79.97.  The dollar continues its downtrend, giving back most of the gains it made last Friday.  Its hard to tell these days if there is much a link between the dollar and PM.  Today for instance the dollar and gold moved up together during one period, but did the opposite in other periods.

Mining shares showed some promise today; GDX was up +1.65% on moderate volume, while GDXJ was down -0.47% also on moderate volume.   GDX rallied early, sold off as is the recent pattern, but then all day long it rallied slowly finally closing the day close to the high.  The volume wasn't dramatic, and the price movement wasn't particularly large, but together I'd have to say it forms a cautiously bullish picture, especially closing at the day high.  When was the last time we saw GDX close at its high?  That was 12 trading days ago, after the Fed No Taper announcement.

This was not a safe haven move, or related to the shutdown, or anything like that - in my opinion.  My interpretation was: someone decided to pop silver above its trading range, and then kept buying silver contracts through end of day.  2500 silver contracts in a 1 minute spike is really a huge amount - to put this in perspective, Managed Money only had 12,400 short contracts as per the most recent COT report.  Gold didn't didn't get the same level of love at all; it lost 40% of its gains by end of day.

This whole performance of silver vs gold reminds me a little of how PM was behaving back late July/early August.  Perhaps the miners are the key.  If traders keep buying them the way they did today, perhaps we might have something here.

 

5 Comments

davefairtex's picture
davefairtex
Status: Diamond Member (Offline)
Joined: Sep 3 2008
Posts: 5072
new SGE premiums over COMEX chart

The Shanghai Gold Exchange was closed for a week due to China's National Golden Week holiday.

Today was the first trading day in the SGE since Sep 30.  The current premium over COMEX: +6.58.  Looks moderately bullish; premium actually rising with the price of gold.

Oh, fun website for figuring out such things:  http://www.timeanddate.com/holidays/china/


 

Jim H's picture
Jim H
Status: Diamond Member (Offline)
Joined: Jun 8 2009
Posts: 2379
Question from the conspiracy theorist...

What are these numbers you are quoting as SGE premiums? 

You have called these numbers the premium, in dollars, over Comex Gold price...but the numbers you quote don't make sense when I compare them to the data presented by Koos Jansen, a Gold blogger who focuses heavily on the China Gold market.  Koos is nice enough to show us the charts taken from SGE, with headers in Chinese, on which he bases his data, so I have no reason to question him.  While your historical chart shows the same up's and down's as that of Koos, your premium calculation is clearly different (lower - what a surprise!) and your data actually goes negative in the early part of the year (which explains your past comment about the existence of negative premiums) whereas Koos' does not, other than for a day or two.  If I had to guess, I would say that you are somehow confusing the absolute value of the Yuan premium per gram (the numbers in the last column of the chart below) with the dollar premium per ounce.  Koos' chart is nice in that he expresses the premium in hard to confuse units of %, hence a 1% SGE premium (over the international Gold price set by Comex) is clearly about $13 and change, and anyone with eyeballs can see that the premium has been centered around 1% for the last month pictured in the graph.  

Premiums based on the SGE week reports. Difference between SGE gold price in yuan and international gold price in yuan.
SGE+premiums.png

  Screen dump of premium section; the first column is the date, the third the international gold price in yuan, the fourth is the SGE price, and the last is the difference.
 

Schermafbeelding+2013-10-04+om+10.27.44.png

link:  http://koosjansen.blogspot.com/2013/10/week-39-shanghai-gold-exchange-ph...

davefairtex's picture
davefairtex
Status: Diamond Member (Offline)
Joined: Sep 3 2008
Posts: 5072
shanghai premium calculation

Here's how I figure out the shanghai premiums.  Its a bit complicated, but I try hard to get an accurate number.  I'll go step by step through the process for yesterday's value (2013-10-08):

Every day I get minute-by-minute OHLC prices from the current active contract on COMEX.  For a given day, I extract the COMEX closing price from the 0330 EST tick, which (I believe) is the shanghai close time.  Yesterday 2013-10-08 at 0330 EST, the front month contract (GCZ3) gold was 1327.10.

I get my Shanghai close prices from the website sge.sh: here's an example quote page from yesterday:  http://www.sge.sh/publish/sgeen/sge_price/sge_price_daily/10586.htm

I pick the AU(T+D) item, and select the closing price.  Yesterday, that value was 263.06.

Then I get the closing price for USD.CNY.  Yesterday, it was 0.1630.

Then I use a constant to convert grams to troy ounces, that's 31.1034768.

The final equation is: 

FX.CNY.USD * SGE.AuTD.Close * 31.1034768 - COMEX.Gold.SGEClose

Values for yesterday are:

0.1630 * 263.06 * 31.1034768 - 1327.10 => $6.55

One place that I might have an error is, that Close column really isn't set at 0330 EST, it is some other time.  It occurs to me at this moment that ... this might have a problem when US moves into Standard Time.  I've only been doing this during EDT.  Another is my selection of the instrument in Shanghai - I picked the Au(T+D) item because it had a value in the "delivery volume" column which I took to mean it was the one actually delivering gold bars.  It is possible your source chose a different instrument.

I might be pulling the wrong tick - perhaps I should get the close price from the 0329 tick instead.

Its also possible your source uses a different value for the "international price" for gold - it might be the LBMA AM fix, the LBMA PM Fix, some other spot price I don't know about, or the COMEX closing price rather than the COMEX price at the moment Shanghai closed.  Might be worth asking the gentleman about.

If you find where I've made the wrong assumption, or you think some other instrument would better represent "the price of physical gold in shanghai vs COMEX" please let me know.

Jim H's picture
Jim H
Status: Diamond Member (Offline)
Joined: Jun 8 2009
Posts: 2379
SGE premium

In a quick review I don't know where the difference lies, but I am pretty sure that your number would not tend to be systematically low by such a large factor if it were just a choice of closing price time.. as this would tend to introduce a more random type of difference, vs. a constant bias like we are seeing.  Does this make sense?  

 

davefairtex's picture
davefairtex
Status: Diamond Member (Offline)
Joined: Sep 3 2008
Posts: 5072
more detail on shanghai

Yes I agree Jim, it seems to be a systematically lower difference.  But I may have something that explains it.

Jesse has some details on the SGE:

http://jessescrossroadscafe.blogspot.com/2013/08/some-local-color-on-shanghai-gold.html

Here's the important bits:

I don't know if you are familiar with the operation of Shanghai Gold Exchange. The most frequently traded gold products are Au9999, Au9995 and Au(T+D). The most frequently traded silver product is Ag(T+D).

Au9999 and Au9995 are cash products. Au9999 is 1 kg of gold 99.99% pure and Au9995 is 1 kg of gold 99.95% pure. You just trade them like trading stocks.

Au (T+D) (the contract size is 1kg of gold no less than 99.95% pure) is similar to futures contracts but it does NOT have a specific delivery month. Delivery can take place every trading day. Between 15:00-15:30 (Beijing time, GMT+8) of every trading day, all the longs/shorts can submit their intentions to take/make delivery.

If there are more intentions to take delivery than those to make delivery, all the shorts of Au (T+D) must pay the longs Deferred Compensation Fee: short to long. If there are more intentions to make delivery than those to take delivery, all the longs of Au (T+D) must pay all the shorts Deferred Compensation Fee: long to short.

So, I think I picked the right instrument when I selected Au(T+D).  And it appears that your source chooses the cash-settled contract instead of the one that involves delivery.  Specifically, for the day 2013-09-27:

http://www.sge.sh/publish/sgeen/sge_price/sge_price_daily/10574.htm

You can see he uses the Au9999 cash-settled contract, since he uses 263.58 in the SGE price column, which matches with the closing price for Au9999, while Au(T+D) was 262.89.  I'm not sure why Au(T+D) is priced lower than the Au9999 contract, but that might account for the difference you spotted.

If you really want to get a gold bar, you can buy one of those Au(T+D) contracts, and demand delivery each day.  If there aren't enough shorts to supply the longs, then you get paid the penalty fee.  I wonder how it all works out in practice.  No doubt that penalty is responsible for the lower price for Au(T+D).  You have to really be serious about either supplying gold (as a short) or taking delivery (as a long).

EDIT: Did a bit more research.  Penalty fee isn't all that large: 0.02% per day, which amounts to 7.3% annually.  Not the end of the world, but becomes annoying over time.  For a kilo bar, its about $8/day.

http://www.cib.com.cn/netbank/en/Personal_Banking/RMB_Wealth_Management/Personal_Precious_Metals_Trading.html

Thanks Jim, without your question, I would never have run this down.
 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Login or Register to post comments