Serving the Quantitative Finance Community

 
User avatar
sslack88
Topic Author
Posts: 1
Joined: August 22nd, 2005, 8:38 pm

Reconstructing order book from full depth data

March 29th, 2019, 9:24 pm

I'm working on a research paper and I have some full depth tick data for several ETF's.  I need the ability to see what the exact bid/ask spread was at any given moment, so essentially I need to reconstruct the order book.  I've found a few reference papers on the subject, but they weren't very useful.  Could anyone give me some useful guidance/steps on how I can achieve the result I'm after?  I've attached a screen shot of a data sample.  Once I understand what I need to do, I can write the program to process the data.
Attachments
full-depth-data.png
 
PrinceQnt
Posts: 9
Joined: October 27th, 2020, 11:36 pm

Re: Reconstructing order book from full depth data

December 23rd, 2020, 1:12 pm

Have done some of this modeling in the past on Futures. Did not find any papers of interest, just a rule set based on exchange rules, with equities it can become more challenging given the fragmentation. 

But really if you are interested in the spread only, and are willing to assume what was shown was "real", regardless of size, its just best offer best bid. If they cross/hit/lift without a cancel event away from that price and bid/offer does not re-bid/re-offer then you lean on last un removed best bid or offer. This of course is reconciled with trades book @Time. 

If the source is timestamped/coloed well, things should line up quite well with simple rules. 

Of course if you have some more specific questions, post I will try my best to help.