Page 1 of 1

how do we estimate position of our order in order book?

Posted: July 10th, 2012, 10:51 pm
by saminny
how would you estimate your order in the exchange order book? The order of order events and acks is not deterministic or guaranteed. How could you write an algo to estimate accurately your position in the order queue? Note that when we send an order to exchange, we get acknowledgement back from exchange, followed by market data add order event for our order. At the same time, we will receive order events for other orders. The order of these is not guaranteed. The goal is to estimate our position in the order queue.I know its exchange dependent but I just need some ideas on this topic. Thanks

how do we estimate position of our order in order book?

Posted: July 24th, 2012, 2:05 am
by winstontj
Send odd-lots and bring up the L2 on screen??? (kidding)Assuming you have true DMA it's fairly easy if you have decent time synchronizations and really, truly know your latency (and can calculate it all very fast) as well as subscribe directly to the exchange data feeds. If you subscribe directly to the exchange feeds you will know how much volume there is at a certain price. From there you look at the timestamps and do the math. As outrun (below) points out it isn't an exact science but over time you can average out and find out where you jump in at a certain price. It takes a bit of a correction factor (you will learn where your estimate puts you) but you can tell fairly accurately over time.

how do we estimate position of our order in order book?

Posted: July 29th, 2012, 7:53 pm
by tu160
QuoteOriginally posted by: winstontjSend odd-lots and bring up the L2 on screen??? (kidding)Assuming you have true DMA it's fairly easy if you have decent time synchronizations and really, truly know your latency (and can calculate it all very fast) as well as subscribe directly to the exchange data feeds. If you subscribe directly to the exchange feeds you will know how much volume there is at a certain price. From there you look at the timestamps and do the math. Do you still need a model for cancellations of orders standing in front of you with parameters to calibrate? Or there are data feeds to get this info today?