This is the spread built from trading a block of iShare's Silver Trust (SLV) shares against one CME-listed large Silver (SI) contract. The December, 2011 silver contract was recorded in this study due to the low bid/ask spread and subsequent liquidity. The SI contract begins trading on Sunday evening, hours before liquidity is introduced to the SLV shares, visible on the left of Chart A. The introduction of SLV liquidity allows for the construction of the spread.
There was a slight break that occurred during the opening minute of SLV, however the deviation appears to have been caused by halted quotes in the SI leg of the trade. It remains to be seen whether this halt in SI was a error on my end on account of my taq data collection software or if the SI contract actually was halted at the exchange level for several seconds. Below are the corresponding prices of SLV and SI (left) and the deviation in the spread (right).
September 27, 2011
September 26 - 27, 2011
There was a slight break that occurred during the opening minute of SLV, however the deviation appears to have been caused by halted quotes in the SI leg of the trade. It remains to be seen whether this halt in SI was a error on my end on account of my taq data collection software or if the SI contract actually was halted at the exchange level for several seconds. Below are the corresponding prices of SLV and SI (left) and the deviation in the spread (right).
Level 1 Volumes & Prices During the Opening Minute
Below are the level 1 bid/ask prices and corresponding volumes for the two securities. The offered volume on SLV (the red bars) began to increase as the prices of SLV rallied while the SI contract, for whatever reason, remained unchanged. This appears to be evidence of high-frequency firms placing resting orders on the SLV offer in an attempt to capture this mispricing. Around 08:30:53, prices in SI revert to 'normalcy' and the spread collapses.
![]() |
| Chart D |
Similar, although less severe, breaks did occur during both open markets (see Chart E.1), however these will be explored in forthcoming posts. The volume of quote changes makes this amount of data very difficult and tedious to work through. Excel becomes very unstable and unresponsive operating at the "zero bound". If anyone has any recommendations on alleviating this issue, I am all ears.
September 27, 2011
![]() |
| Chart E |
![]() |
| Chart E.1 |










3 comments:
Sorry this comment is a little late, but I use R for this type of thing, much more flexible. But you will need a good amount of memory.
Are you actually trading these deviations?
Hi Craig, thanks for visiting. I would be very interested to hear what else you're working on in the markets if you are open to email dialogue.
I currently do not trade the strategies I post on FRTD. I have neither the coding acumen or the technology budget to properly implement the strategies on here. I publish some of the most basic spreading strategies while keeping many of them private.
I currently trade discretionary spreads across the commodity complex.
Sure, drop me an email if you want.
I'm just trading basic equity stat-arb type of stuff, but I'm always keen on swapping ideas.
Post a Comment