You might think that the mean(abs(close-open)) pattern is largely caused by the changes in volume (there are far more zeros at night, which reduce the average), and you'd be right. With these removed, you still see the same diurnal cycle during the day session though (attached)
I guess you could argue this is expected because there are more "fundamental" announcements in the morning, and so doesn't necessarily indicate market inefficiency??
There is nothing unusual at all in that pattern. Thats how market participants trade. I wouldn't say it has to do with "fundamental announcements". All instruments world wide will display the same volume pattern.
What were you looking for? Did you find it?
But what about the others? E.g. "transactions per unit volume" seems kind of interesting - on average, people put through more units per order at the end of the day than the beginning.
Units you mean contracts? Nope that is standard. Retail and the crap prop have nicked off for the day, most likely stopped out but also a lot just have a few swings and nick off for the day. All that is left is Algos and good Prop traders thus the larger size.
What about the increase in contracts per trade in the morning and afternoon, relative to midday and night session? I was thinking that this is when the largest fraction of the trades is probably by humans, and so maybe humans do more contracts per trade than bots.
, whether you are a bot being designed by a person or a fool trying to beat the former.
in fact the 5th/95th percentiles from 2011 lie pretty much on top of the 1st/99th percentiles from 2012, so yeah, a massive reduction in volatility or range or whatever.
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?