MetalsTrades Commentary - August 5th 2012
The past volatile trading week encompassed the FOMC ECB and NFP events. Now that these events are over, it marks the transit into August summer doldrums. In view of this, I thought it important to look at just how this year’s thin summer market trading will likely differ from all years past. Net of algo’s, market depth on Thursday and Friday was the thinnest all year with a lot of desks now manned by junior’s, however, there was no doubt all hands were on deck for the FOMC, ECB and Fridays NFP. Once defence of positions and chart painting was complete, these desks returned control back to the juniors marking the point when full on summer trading commences.
As expected, as soon as we exited July rollover trading, market depth fell off a cliff. This is not to be confused with trading volumes which are bolstered by the increased appearance of algo based trading platforms. Prior to 2010, algo’s were not much of factor during thin August summer trading. That is not to say we didn’t have volatility but it was fairly measurable within clearly defined medium term ranges. This is no longer the case as this type of computerised trading is exponentially increasing.
In last week’s commentary, we took a look at how the Exchanges, (specifically the CME), are losing clients and a resulting decline in open interest. The bulk of this decline was caused by the MFG collapse and the losses customers incurred as a direct result of the CME failing their primary fiduciary duty in protecting client moneys. Since then, (after PFG collapse), the CME wrote and published letters trying to reassure clients, attempting to draw them back to do business on the exchange. However, this letter fell grossly short of its objective as it had paid no regard to hundreds of other complaints from non-exchange members regarding the rapid increase of HFT volumes that were distorting price discovery and creating unnecessary volatility, especially noticeable in the smaller markets such as silver. As an example, on Wednesday, open interest plunged 6,625 lots, 20.16 tonnes or 1.64%, to 396,778 lots. This is the lowest open interest since April 24th and the second lowest this year. Except for 5 days in April, open interest has not been below 400,000 lots since September 1st 2009. On Thursday, OI declined even further to 394.877.
Being focused solely on the fees, the CME has continued to gear up very large investments in HFT platforms to accommodate member firms. The net result of this has and continues to be a secondary reason for an exodus of real money from the exchange, primarily leaving algo’s trolling algo’s but in much thinner conditions than we have experienced up until now. As stated in commentary last week, I believe the CME has now caught themselves between a rock and a hard place.
This brings us to how this summer’s doldrums will almost certainly differ from prior years. Neutral algo’s are not the real issue here as they don’t so much focus on direction as taking a little of both sides of real market bids and offers. Whether they provide any value or not is another question.
The issue here is the malware employed by 2 COT’s. These 2 Bullion Banks hold the book on 98% of all gold and silver derivatives and as a result know where the stops and margin positions are on both sides of the much larger OTC market, also knowing these positions are mirrored in the Comex. Through size, leverage and concentration of short positions on the Comex, it is these directional algo’s that are specifically deigned to lead the many ‘neutral algo’s’ by the nose, providing the necessary additional volumes to manufacture downside waterfalls that are specifically targeting vulnerable and visible long stops held by MM and specs but also large OTC positions. This added volume also helps to hide their footprints as currently the illegal practice of ‘spoofing’, (Disproportionate Bids and Offers that are momentarily placed into the market but not intended to trade, creating exploitable air pockets under the market), even though the data is available, it is not yet disclosed by the CME whose members for obvious reasons do not wish this level of transparency exposing their footprints.
We regularly identify their signature footprints, usually first spoofing, which in picoseconds alerts neutral algo’s that bid interest is evaporating, then as market makers fully activating their proprietary malware, hitting/ pulling bids ahead or upon momentary uncertainty around pivots or as most recently around news releases that would normally be minor events. QE on/off happens to be the current lever. This activity will be much more apparent during this summer’s doldrums which now that NFP traders have returned back to the Hamptons etc. and juniors seek to try to make a name for themselves on the desks, will really commence. These moves will mean nothing fundamentally and due to the very strong physical markets now clashing with the declining 50 DMA sell trigger, will be short lived, but I advise anyone with margined positions to take some very short term cheap low delta put insurance for the month.
A rare clue as to just how much higher sovereign and CB bids have risen during July is highlighted in the following mainstream pieces and gives away a tiny piece of the picture in what we have been noting daily at the fixes.
One of the items that made mainstream news this week that were of note was South Korea's central bank reported it bought an additional 16 tonnes of gold in July on the open market. It is not the size of the purchases, which is quite small in comparison to what we have been reporting in the physical market, but what was illustrative of what we have been observing was the statement made by the CB … “Lee declined to provide the exact net purchase price per ounce it paid for the bullion, typical of most central banks, although he said price was not a major factor in the decision to make the purchases.” Calculations from the total value and volume of the purchase show that the Bank of Korea paid about $1,582 per ounce on average, compared with an average spot price of about $1,592 for the month. The key thing to take away from this is the statement that price was not a major factor in the decision to make the purchases. Not only did we report seeing CB and sovereign buying in this area but in much larger size both in $ and much larger in EUR terms.
Then UBS observed,
“What's perhaps more interesting in this news is the price level which the official sector have recently been buying gold. The biggest reported month for gold purchases this year was March, when central banks bought 76.5 tonnes, and this activity prevented gold from a sub-$1600 move. Again in May, when the official sector purchased 19.3 tonnes this buying inserted a price floor at $1530. Since May the acceptable price level has clearly risen, with the average price in July at $1595. Central bank activity is in part filling in for the tame retail physical buying from the likes of India of late by helping gold on the downside and inserting a price floor.”
It is not usual for this type of news to make the mainstream, but bear in mind these are the reported purchases and do not reflect the full picture which we observe as many times larger.
How this relates to current trading is summed up in a post from Thursday night as we closed near the 1586.30 low … “After covering into the fix COTs clearly made a move to close the pit just below the 50 DMA, this is no coincidence as it is designed to get the hot money short on rallies. I do not see enough mileage to be gained. 1585 /90 is now a CB buy area, (more on this in commentary on Sunday), and am doubtful COT’s are seeking to sell this level into strong physical demand.. This is no more than usual COT NFP gaming but in very thin conditions. No technical damage done…..”
In other words, COT’s are pinned between CB bids and how much they can milk from already very short weak hands. COT’s rang the register on MM shorts on Friday into NFP but then attempted to draw them short again but ran into the same strong physical buyers who we had noted had spot indexed below 1600 and would be seeking allocation into the fix. This happened in very large size. Yet again, these weak hands have been run ragged as captured in the COT report and seen into Friday.
The weekly COT report is as usual little more than a stale dated tool for gaming the weaker hands who base their next week’s positioning on this rearward looking data. The bigger picture is still extremely bullish despite the large Commercial short additions and long capitulations captured in the Tuesday’s report cut off. The 19,800 net short additions in GC were based upon a low base and notably our usual suspects added very little of these positions, in fact since Tuesday, we saw them very actively shorting then ringing the register on those drawn in. Footprints since the report evidence them fully covered on these new additions by Friday’s close. In silver, our usual single suspect added over 3K shorts; this action was 100% defensive and mirrored what they have been doing in Gold for weeks now, as they sought additional mileage in keeping the MM & specs, selling rally’s in knee jerk fashion as long as price stay’s under the 50 DMA. Much like gold, this average has now declined into a very strong Physical buy area and is not sustainable based upon real market conditions.
I will update all levels on Monday.