Not entirely sure if you're trolling. Plenty of red herrings in how you describe the CME and margin rules for participants. You've come up with lots of assumptions without having access to historical tick data, historical depth of market, etc. If I found a "misbehaving" algo in an illiquid product / out of hours the very last thing I'd do is post it on ET. Your observations about there being only a few entities quoting in the overnight session (and therefore likely getting flow information based on how they are filled) may be valid, but I don't see how this would translate into a profitable strategy for you.
My chart doesn't work well on this day, but this is last Thursday- Thanksgiving- and look to the left over 15 minutes from 12:39PM ET - 12:54PM ET. 3,400 contracts down 50 cents then 3,400 contracts right back up 50 cents. I was active in the market then and this was a classic "hole" (I call it "an HFT digging a hole"):
The CME's relationships with its "liquidity providers" are open and are exactly as I described. I'm not here to focus on profitable trading. Nothing I do is replicable. Rather, I will help anyone who wants to trade the CME to understand who they are trading "against." Only what I know to be true. No speculation. Essentially every trade made by hand on the CME is with a machine as the counterparty. UBS fired all of its commodity traders and is replacing them with machines.
Here was just a price-"defending" attack. This one pushed exactly 25 cents. These attacks always have an exact target price push, in increments of 5 cents, from 10 to 30 cents. 10 and 15 cents are the most common. These attacks are to attain their target and there is no meaningful constraint on size (here, 1400 contracts)- they stop exactly when they hit the target, all within 2 seconds at most. Reminding, this is not "buying" like a rational person would do, since this buyer is paying the highest price possible, not the lowest. This is repricing- an agenda with almost infinite size.
Just had another "hole"- Crude a nice value hold after collapsing $1 right after the spike above. So an attack at 11:59:35 of 15 cents down (280 contracts) then 15 cents up after 45 seconds flushed out any weak-hands and/or robots. Trading this, the "signature" is the key- the single burst suggests a machine, not any sentiment or conviction among retail traders. My single recommendation to any daytrader is to make sure you consider Volume as well as Price. PS Note a smaller 10-cent attack right before, at 11:59:15, with classic signatures of the doubled-orders (what I call a "double-barrel blast"). The orders are doubled to stop out any robots who instantaneously fade the first move- I'm guessing most guys now have their robots wait before fading a move (I wait 30 secs).
Looked at two of your examples using volume at price. It appears that most volume tends to be concentrated at the peak of the move with 'path' leading to it being mostly low volume. However, I have to warn everybody that this 'signature' is typical for many moves, not necessarily just those suggested by None Business. Many continuation moves start like this too. I guess he developed a feel or other quantification of the move to be able to fade it (assuming he does so). One thing I agree with him 100% though is that volume is extremely important...
I'm being very careful. The first "move" had no buildup at all and the second was immediately undone, making it not a "move" at all. (same guys didn't suddenly change their mind). Anyway, this isn't going over well so I will leave you guys to it.
It's unfortunate if my remarks unsettle you that much. I stand by my statements about that signature being typical for many continuation moves as well as reversals (that, too, unfortunately-if every move with this signature could be faded I would be a billionaire by now). Still find your observations interesting, no need to take offense and stop here just because someone is not taking it blindly and actually thinking about it.
I think these moves need to be isolated down to the "context" that they appear in. Why would a trader be willing to throw though size of trades in at the market when by splitting them to reduce market impact? Now I might not be the smartest tool in the box but I can think of two reasons. 1. An area of stop losses gets triggered with enough size to consume all the liquidity in the book until they find resting limits to fill them. -Check out Episodic Volatility. 2. A machine knows the markets quiet can trade with itself to move price and has an idea of open interest with resting stops in a nearby vicinity. There is a reason why traders try to hide size and a reason why they display it. Thats why it all comes down to "context". My 2 cents.
None Business, I appreciate your posts immensely. Please continue whenever you feel you have something you can share. Do not be concerned about whether any of us find value in it. It's unnecessary to respond to anything except that which furthers the directions you're personally interested in. I hope this helps you understand at least one view of the value you add to this forum.