Thanks to an alert reader for passing along this note from Market Delta regarding changes in how the CME is reporting volume data. By unbundling trades and reporting multiple, smaller transactions where only one large trade was formerly reported, the exchange has made it more difficult to track the activity of large traders.
I'll be looking into this in the near future. Depending upon the unbundling scheme, it may be possible to still monitor large traders simply by lowering the threshhold for what qualifies as a large trade. If the trades are unbundled into one and two lots, of course, this would be impossible. In that event, it would be necessary to aggregate the trade data over short time periods to infer the presence and activity of large participants.
More to come.
.
Subscribe to:
Post Comments (Atom)
Posts
4 comments:
Isn't this old news. The big brokers compete on who has the best algorithms to hide big blocks of trades.
Yes, and this is why I've always been weary of filtering for large trades. You can often get a completely wrong picture due to all the broken up trades that one thinks are small lots.
Dr. Brett, I'm really interested to hear if you can find some solutions to this issue. I believe there is software out there that is reverse engineering all of the iceberg orders and algorithms and putting back all of the broken up trades together to show them as the original big lots. If I'm not mistaken, I remember reading that the guy at camron systems that you once referenced has created software that does that. But he was trading the Aussie markets. I wonder if there are others providing this kind of software! I'm sure if there are, your connections at all the prop firms and banks could lead you to it.
Dr. Brett, I certainly hope you investigate this further and keep us posted. For those of us that rely on the tape, this may turn out to be a huge problem. I just don't understand why the exchanges seem to be constanlty working against the little guy.
I think the exchanges are in the pockets of GS and the like and who knows who they work for, besides themselves.
I have noticed over a 3 to 4 year period that filtering delta has at times worked well in the ES and at other times not so. There are lots of possible explanations for this but the time and sales bid offer data we receive is often suspect imo. The TF was an extreme example of this for a few months after it swopped exchanges when even the non filtered total cumulative delta was up the swanny.
I shall be very interested to see any further information on this latest change.
Post a Comment