IG asked me to show some screencopies of PRT forms. AT explaning those, I think I could reason out that it is the very same issue as what this topic was created for by B&L.
- A Limit is taken just (very) near to the end of the bar (1 minute in my case) – say a few milliseconds.
- The broker sends feedback, but this happens in the new bar;
- PRT kills de Limit (mind you, which I see still there in the new bar !) because in the bar what the Limit was meant for, no position was taken (that is, PRT does not know this yet);
- PRT thus kills something which already has lead to a filled order;
- IG does not understand a hoot of it and closes the position right away (at a loss and the loss depends on the position size and on the amount the price moved meanwhile.
The only luck we have is that PRT does not kill the System for this reason (thus the system keeps on running).
N.b.: I could work this out because I could see (by guarantee) that the order was filled in the next new bar of where the Limit was put in. In my first description from yesterday this was in minute 03 where it should have been in minute 02 (minutes numbers show above the bars). Below you can see this happening per means of the Backtest (blue arrow) which executes this as should – in the proper bar. Further data also tells me that the Limit order for minute 02 in live was executed in minute 03.
<hr />
So who shall we blame now, knowing this new data ?
What about some lacking order type towards the broker ?
“Hey, I am a Limit order which wants to be cancelled and for which a position exists in the mean time – kill me please !”
Only with this, IG will not see it as some anomaly or improper order from PRT.
1 user thanked author for this post.