Here's how it ought to work ...
Assuming I see a normal open high and drop, with weak but slowly improving buy percentages, I try to guess if it looks like it will do a normal reversion to the mean.
If so, I guesstimate the intra-day spread, a low, what the reversion to the mean would be, etc.
Today had several things looking "right".
_Time PerVol Per Min Per Max _Per $ Val _Per VWAP
10:25 55,335 $1.4100 $1.4968 $79,422.50 $1.435303
11:58 65,826 $1.3800 $1.4200 $92,251.13 $1.401439
12:17 12,310 $1.3990 $1.4100 $17,241.80 $1.400634
Then buy percentages are important. If they start very low and I believe they'll have a semi-normal recovery into the mid-3x% or better as MMs and/or shorters take advantage of the low covering buy prices, the trade starts to look "interesting".
09:30: buy 00.00%, sell 100.0%, unk 0.00%
09:45: buy 16.91%, sell 78.10%, unk 4.99%
10:00: buy 25.70%, sell 61.35%, unk 12.95%
10:15: buy 23.44%, sell 66.89%, unk 9.67%
10:30: buy 23.03%, sell 67.36%, unk 9.60%
10:45: buy 24.99%, sell 65.71%, unk 9.30%
11:00: buy 29.57%, sell 60.61%, unk 9.82%
My buys finished filling at, last trade, 10:52.
So things were going as I wanted to see, generally, and the risk was looking "well reasoned" in retrospect.
Note my buy price was well below the VWAP through 10:30 (last trade in that period was 10:25) and is still below the subsequent period VWAPs.
Now I have to rely on the MMs and/or shorters "doing their job", bringing about a few periods of higher prices as they make $ by both first sell high and then buy low and garnering fees from the exchanges for providing liquidity.
Their gain depends on both that buy/sell mantra and volume and they'll move the price around as they see fit, when retail doesn't overwhelm their ability - and that is the case today and most days recently - to accomplish that, likely taking advantage of day traders that are programmatic in nature (buy at the close sell at the open or vice-versa) and/or momo traders (not likely today) and/or shorters that are covering.
That's my theory and I'm sticking to it! )
Of course, much of the data I show above is not available at the time I need to decide. So that's where the old SWAG (Scientific Wild-Assed Guess), a.k.a. "gut hunch" comes in.
This latter facility has been refined over time by many cuts, bruisings, self-flagellations, etc.
It's reliability is highly suspect, so great care, many deep breaths and small trading blocks are prudent.
MHO, in partial ignorance and always incomplete information,
Bill