Registered: Sep 2012
10-18-12 01:35 PM
Quote from tfc_trad3r:
Once you trashed the Bloomberg terminal for their data quality, your ignorance was apparent. You want to test on data that is as close to raw as possible.
Quants want to run their simulations on data that reflects the true trading session. By changing the order of the data, Quant Quote is not a true representation of the session but an interpretation of the actual events. The quant quote scrubbing / filtering process is an example of revisionist history.
Yea...you should get your eyes checked. I did not say there was anything wrong with the Bloomberg terminal. I used to work at a firm that had access to Bloomberg, if you look at the terminal closely, you will see that out of sequence trades are quite common. That was my only point regarding Bloomberg.
QuantQuote also offers Tick resolution data:
If you look at their Tick resolution data docs, you will see that they do NO FILTERING, and include every single trade, including out of sequence trades, along with their sale condition.
However, as I demonstrated in my earlier example, QuantQuote's methodology of removing late reported trades is the ONLY unbiased way to represent the data. Did you read the QuantQuote FAQ I linked to earlier? (https://quantquote.com/support_glossary.php#OFS)
QuantQuote is NOT changing the order of the data, they just don't use late reported trades when calculating low/high for a minute bar. That is the MOST accurate way to represent the data in a one minute bar format.
At the firm I used to work at, when reading off of live feeds, we also threw out trades with sale condition Z (out of sequence) for the exact reasons I brought up earlier so believe it or not, this is very common practice among truly professional quants.
Now, if you still have any doubts about the validity of this process, note that Tickdata (ridiculously expensive, but also well regarded) also performs this same type of filtering. Look at the second white paper here: http://www.tickdata.com/support/white-papers/