terminal exit order

[NinjaTrader_Ray]

To explain anything, I would need a simplified strategy that demonstrates this new limitation that works in backtest but not in real-time. Logic should be the same in both situations and I would need to investigate this further.

[nfeldberg]

OK, played around and got some progress, but got another wall to workaround.

I figured out that I do not need to cancel the exit orders before trying to flatten the position in prep for reversal, so the resubmit of the exit/stop is not a problem any more....

however, I am getting the same error when I am trying to close the open position with a limit order. (first try to close work fine, however if the first try did not executed (i.e. price did not match), the exit order is cancelled at the end of the bar, than if another reversal signal happens, and I try to close the position I am getting the terminal exit order already....

again it happns only with the sim and not with backtest.

if you could explain the internal logic that you are using to not realy cancel the orders it might help to find a workaround, market monitoring is not an option.

does the same problem will happen if I am flat trying to go into a position with limit order, price did not match, order cancelled, than trying to go again into position and getting this error??? if yes it is a real basic functionality limitation which means it is impossible to work with entry limit orders in your platform.

maybe I am naive, but I spent too much time in learning this platform and migrating my strategy from TS, to declare a defeat, please explain what the logic is, it might help to find a workaround.

thx Nir

[NinjaTrader_Ray]

As you have asked --> Yes.

However, based on what I know --> No, since what you can't do is re-submit the cancelled exit order.

What you likely could is when your reverse condition is true, monitor the market to see if the limit price is touched, then cancel the stop, wait for confirmation and send market order to enter.

I can't guarantee that this gets around the problem but it would be another approach to try.

[nfeldberg]

I have multiple entry and exits (i.e. 1,2,3,4) to enable scale out.

if I will not reference the entry the exit will not work properly.

so to make things clear let me ask a question again:

is it possible with ninja to get into position, send 1 stop and several limits (to enable scale out) and than as required to reverse via a limit order when you have an opposite signal?

by now I am already an "advanced programmer" on your platform so help me by providing an approach, that would not hit a wall with hidden internal logic of the current version (which is not documented and so far not explained).

thx,
Nir

[NinjaTrader_Ray]

When submitting an entry, do not pass in a signal name, just leave it blank, meaning, emptry string.

EnterLong() vs EnterLong("mySignal"), same goes for exits.

I can guarantee that this will work, just throwing out an idea that may bypass the road block that you have encountered.

I surely understand the disapointment of the effort involved only to find out you hit a wall. Unfortunately, this limitation is not a trivial item to change and thus its on our NT7 list.

I have provided an idea that may work, if not, you will have to take into account this limitation and see if you are able to work around it.

[nfeldberg]

thx ray,

I am not sure I follow your suggestion regarding the empty string for which order the original or the new one?, if I would change the name would it work?

now this limitation is very disappointing, I would like to refer you to the following thread, I have followed your suggestion on how to achieve this rather simple functionality, and after mastering your advance programming it looks like your suggested solution is not possible. I would appreciate it if you could help me and let me know what the best way to proceed.

http://www.ninjatrader-support.com/vb/showthread.php?t=6007

thx
nir

[NinjaTrader_Ray]

Just thinking...

If possible, use empty string as signal name and see if that circumvents the limitation?

[NinjaTrader_Ray]

We just took a quick look and what you are running into is current expected behavior. We will review this with NT7 as we do believe it should be changed.

The problem (internal to NT) is:

- You submit an exit order
- You cancel
- NT has internal tracking of when to ignore orders
- You resubmit and we currently then ignore it (too complex to explain why)

With this known limitation, you will have to find a work around or different approach with NT 6.5.

[nfeldberg]

it is 100% reproducible, regarding a simplified strategy, that will take some time, as the fact that I have to manage everything by the code (to avoid the order rules) means that it becomes very complicated.

it would be great if Jose can have a look on the logs before I am working on simplification... I would be happy to share the strategy fully with Josh directly if required.

reagrds,
Nir

[NinjaTrader_Ray]

Josh will follow up with you later today. Is this 100% reproducible and can you provide us with a simplified strategy to demonstrate this behaviour?

[nfeldberg]

Hi all,

I am having a problem with resending exit limit and stop orders (they are cancelled to allow reverse and than resent if the reverse was not successful). I am doing it to follow the order handling rules.

my orders are ignored with an error:There already is a matching, terminal exit order in place'

I have checked and I am monitoring and resending the orders only if their status previously has been cancelled.

when I am back testing there is no problem, just on the simulator. (real time)

I am attaching all the logs, any insights would be great.

thx
Nir