ExitShortStop/ExitLongStop problem

[marotom]

Hello NT support!


I have a big critical problem. I'm using the ExitShortStop/ExitLongStop method, in this syntax: ExitLongStop(0, true, quantity, stopPrice, "StopLossLong", "");
every time, when I get sign it will amend at FirstTickOfBar, usually in every 10 min. For a little while it works as expect and randomly from an undeterminable point it won't amend the quantity or stopPrice of the same method but it sends every time at the sign a new ExitLongStop with a new stopPrice or quantity, so it will accumulate many Stop order, and can't able to close the whole position. (it will ignore)I attached a picture about it. I used to Market Replay with 500 x speed. what can accour it? the MR has problem, or the method? please help so I can't trust in NT when I want to run it in live.
when I run it again only the same time period, where the problem accoured, won't give this behaviour. it accour when the strategy runs for a long time. 2-3 days.
thanks. Tom

[NinjaTrader_Bertrand]

Tom, what is the TraceOrders output stating at this time then? Would it depend on how fast you replay this specific strategy? If you resubmit the order again using the same name it should amend it as needed. Does this happen only in this specific strategy? On which NT 7 release version are you running here?

[marotom]

Hello Bertrand,

I use the TraceOrder. I'm using 7.0.1000.7 NT version.
yes, your right, if I resubmit the Stop order using the same name it will perfectly amend as needed. if the quantity or/and stopPrice is equal it will certainly ignore.
I recognise two things that can generate this problem. the accoumulate of the Stop orders begin, when I want to decrease the position, and I use ExitLong/ExitShort method and it was ignored. as can be seen on the picture under SignalName=Alap_Rész_Z_L - 7. after that it was submitted 2 times the new Stop order, and won't amend the original Stop order. since there are 3 stop orders, the position close method can cancell. but there was another case, when so many pending exit order exist the position close order can't able to cancell them and there were ignored.
the speed doesen't effect on it.
the second, when I zoom out the plot executions Marker pass through the closed trade session in spite of there aren't associated positions yet. between the trade session the position always Flat. this can be seen on 2. picture. what can throw it?

2. - I modified the exitorders name and there aren't ignore placeorders. So it doesn't throw this behavior of the problem. as can you see on the 3. picture.

3. - another remark to this continuing the 2. point. I found another: this Stop order accumulation begun when this stoporder will ignore. I solved the exits problem so there isn't any exit order ignore only Stop ignore. after that begun the Stop order accumulate, it can be seen on picture 4.
thanks Tom

[NinjaTrader_Bertrand]

Hi Tom, so you're attempting to scale out then here from a position using the same signal names but different qty's? In order to do this please scale in first if using the managed approach and then you should be able to scale out as needed.

[marotom]

Hi Tom, so you're attempting to scale out then here from a position using the same signal names but different qty's?

-yes, after and before the scale out/in I use only one Exit order with the same name. Only two datas will amend: the qty (depend on the scale out and scale in) and the stopPrice (depend on the actual price in FirstTickOfBar). and after the scale out behavior as a new Stop order would submit with same name and it won't amend the exists Stop order. but I'd like only one Stop order! this is my biggest problem. but I have to remark there are 3 places(conditions) to amend the same original Stop order in one direction long/short, but not at the same time!. it can be, for this reason will confused? do you have any idea what can throw it? it is important to solve it!

thanks Tom

[NinjaTrader_Bertrand]

Tom, is there potentially any simplified code example we could run here on Market Replay to understand the scenario better you're running into? To start off it looks like the sequence of events would need to be degugged and the code used simplified until it works exactly like you would expect, then add your other more advanced parts one by one to see when it breaks and causing you the issues seen - from there you could think of ways to resolve.