Splitting orders

[oferns]

1. Split orders into market orders
2. Allow them to be refreshed

This allows us to see aggression & it's ability.

Thanks

[Derek27]

I'd like to give that a +1, sounds like a great idea, but I haven't a clue what it means.

[oferns]

1. It means we can see who may be sitting on the book, holding price at places making the back/lay look weak, before possibly turning price.

2. It also lets us see if back/lay is actually is as healthy as it looks.

In conclusion, we test the real strength of the book.

[Derek27]

1. It means we can see who may be sitting on the book, holding price at places making the back/lay look weak, before possibly turning price.

2. It also lets us see if back/lay is actually is as healthy as it looks.

In conclusion, we test the real strength of the book.

Do you mean if £1000 is sitting at a price, you want to know that it consist of 3 bets (£650, £297, £53)?

[oferns]

That would be cool. Since we are getting data for free. I doubt we will be getting that level of data. However if we are, then I could see that feature being very handy.

We are getting enough to create the volume profile, consisting of people buying & selling at market price.

I want to that to be split and refreshed upon returning to those same prices (after a time we can predetermine), to see aggression or lack of, moment to moment.

[Derek27]

That would be cool. Since we are getting data for free. I doubt we will be getting that level of data. However if we are, then I could see that feature being very handy.

We are getting enough to create the volume profile, consisting of people buying & selling at market price.

I want to that to be split and refreshed upon returning to those same prices (after a time we can predetermine), to see aggression or lack of, moment to moment.

I'm not sure how feasible that would be. If £500 worth of orders are cancelled, I doubt there's any way of knowing if it was one or five orders cancelled and how much of each one. All you know is £500 has been cancelled so it would be impossible of keeping track of the number of orders.

[oferns]

At this moment all I would appreciate is the profile (completed orders) split, so I could see aggression.

However, the idea you suggested are different, but is equally valid. I think we may be conflating the two.

[weemac]

At this moment all I would appreciate is the profile (completed orders) split, so I could see aggression.

That was the very first idea I had when starting out but never found a solution. I doubt it's feasible due to its granularity.

[oferns]

I came across this which I thought was quite interesting..If you can see the original post, you'll see the volume split.

I didn't take long before I found an example...my sandbox spreadsheet shows the vol change allocated to either back or lay depending on the LTP & current prices.

It runs at 20ms and each line is a 1s summary, newest at the top. During one of those secconds, in one of the 20ms packets, an 18k vol change went through and an LTP that was the same as the last know back price (hence the jump in Back Vol), but I was watching and it was definately a lay (as confirmed by the price rise). The Diff col shows the net diff between back vol and lay vol compared to the previous second.

It's not easy. That screen shot would make more sense if I could show more of it, but I'd rather not.