ES Microstructure

I recently added a custom Level 2 indicator to an automated strategy and ran Market Replay against CL and ES and the performance was worse than without Level 2.
 
Quote from Mr_You:

I recently added a custom Level 2 indicator to an automated strategy and ran Market Replay against CL and ES and the performance was worse than without Level 2.

do you have a working hypothesis as to why this was so?
 
There is so much spoofing and order crossing and gaming in terms of the bid / ask - especially off-hours. I would be curious to know why a serious trader would think that there is a tangible and repeatable edge there ?

ES is heavily automated - and that includes quite a bit of arbitrage going on in terms of spread relationships with other instruments. Look at the COT data for spread volume in the ES amongst institutions and asset managers. My point being - you cannot assume that a lifted offer or hit bid is an aggressive liquidity taker who will on his own accord move the market enough where you can race him for a tic or two. That is my honest opinion, fwiw.

http://www.cftc.gov/dea/futures/financial_lf.htm

Spoofing and trickery regarding the bid / ask spread is a reality - ask some electronic fixed income traders about Harris Brumfield and Paul Rotter.
 
Hey guys I noticed that some of the "last traded price"s have 2 extra digits in a slightly different font. Are they just precision average prices off the latest fills? Or do they mean something else?

Just making sure. thanks in advance.
 
Quote from bone:

There is so much spoofing and order crossing and gaming in terms of the bid / ask - especially off-hours. I would be curious to know why a serious trader would think that there is a tangible and repeatable edge there ?

ES is heavily automated - and that includes quite a bit of arbitrage going on in terms of spread relationships with other instruments. Look at the COT data for spread volume in the ES amongst institutions and asset managers. My point being - you cannot assume that a lifted offer or hit bid is an aggressive liquidity taker who will on his own accord move the market enough where you can race him for a tic or two. That is my honest opinion, fwiw.

http://www.cftc.gov/dea/futures/financial_lf.htm

Spoofing and trickery regarding the bid / ask spread is a reality - ask some electronic fixed income traders about Harris Brumfield and Paul Rotter.

I had an indicator built by a student where hidden Volume from the Dome is displayed when there are certain ratios between bid ask. When he got it done, WOW, I was surprised at what goes on that I can't see cause the volume and spoofing. And your right, the average trader would have no use of knowing these numbers as we tested a multitude of searching some type of reoccuring patterns. The indicator did give me another idea where I had an automated method that runs 16-24 hours on many markets, that idea has to do with hidden volume/time. Unlike most of those that teach others about how to trade off the Dome using volume at price levels, I have an opposite approach.
 
Quote from Mr_You:

I recently added a custom Level 2 indicator to an automated strategy and ran Market Replay against CL and ES and the performance was worse than without Level 2.
Do you notice any relationship between CL and ES?
I see a significant move intra day in the CL and the ES will follow.
 
I see exactly the same behavior in KOSPI 200 futures (ES equivalent on South Korea). I did come across several DOM patterns that repeat all the time. Took at bit of coding top get there.

I am using IB as it is the only venue now to trade this market from US. Generally very happy with them, but as we all know they do not send raw data through.

Planning to start collecting statistics on ES in near future. Theoretically the patterns should be the same.
 
Quote from RedDuke:

I see exactly the same behavior in KOSPI 200 futures (ES equivalent on South Korea). I did come across several DOM patterns that repeat all the time. Took at bit of coding top get there.

I am using IB as it is the only venue now to trade this market from US. Generally very happy with them, but as we all know they do not send raw data through.

Planning to start collecting statistics on ES in near future. Theoretically the patterns should be the same.

Feel free to share what you find. Cheers.
 
Quote from hft:
Your(IB's) feed is just wrong. Some trades are missing, some are interleaved incorrectly, and some pricing is just wrong (the locked market at 1427 that you have for a while for example). In that timespan there were hundreds of quote upates (~700 to be exact, between 9:30:22 and 9:30:24, and about 100 separate trade prints), and even the snippet that you post is obviously missing quite a few. Perhaps the most glaring inaccuracy I see in that screenshot is that the trades aren't aligned with the quote updates properly (that whole clip of trades, from what I can gather, should be shifted up about 10 lines to get in the ballpark). I have raw data captured from a stream connected directly to CME so I can see precision out to nanoseconds. So, long story short, your data is inaccurate, so don't draw any microstructure conclusions from it.
@gmst: Have you reached a conclusion: Is the data from IB wrong, or are you displaying it wrong? How are you displaying it?
 
Quote from rwk:

@gmst: Have you reached a conclusion: Is the data from IB wrong, or are you displaying it wrong? How are you displaying it?

I have not compared IB data with a true feed like IQ FEED or bloomberg. Primarily, I trade at 5 minute frequency, kind of few trades a day - so it doesn't matter to me if the ticks I get is not numbered properly.

In future, I would go ahead and do a proper research on this aspect. Thanks for bringing it back to my memory. These days I am very busy with few other projects.

However, I think HFT has made a strong claim here. Someone from ET should come forward and test this claim by comparing IB feed with some other true feed (bloomberg would be best) and report their findings. If no one comes forward, in any case, I will do a proper comparision but that may be a few months down the line.
 
Back
Top