Different results in Strategy Optimizer, Replay Mode and Demo Trading (Live Data)

[arne]

Hi,

I have implemented a very simple two bar strategy, which places a limit order for entry and a stop loss order both on the same bar. When the limit entry order gets filled, the stop loss order will be trailed until it gets filled too on every new bar close. If the entry signal gets invalid, the limit orders are going to be removed. This works in general.

When I am testing this
- with historical data in the Strategy Optimizer for two days on 1 minute data (dxFeed, RTH, generate Ticks on) the performance is negative (-$248.03).
- with historical data in Replay Mode for the same two day on 1 minute data the performance is positive ($2,050).
- with live market data applied on a chart window and activated strategy, the orders are not getting placed. (tested with NQ, ES, RTY and BTC Futures)

Am I missing something here?
Why is it working and placing orders in Strategy Optimizer and Replay Mode, but not in a chart in live trading environment?
And why is the performance different between Optimizer and Replay Mode with the same data?

I am using the latest professional edition of MW 6.9.0 on macOS with Java 21.0.2 (64bit).

Thanks for your help.

Best,
Arne

 

[MrSharky]

I'm getting strange Replay results on simple bracket orders with preset stop and profit targets. I think there is a bug somewhere but nothing confirmed from motivewave.

Replay potential Bug

I have been playing around with the Replay function (testing a strategy) and come across some worrying discoveries. A bracket order with fixed stop and profit limit, say as a example 200 USD Stop and a 400 USD profit LMT order. The stop will execute at the 200 USD threshold but the profit will...

support.motivewave.com

 

[Aionda]

I also get 3 different results.
In normal replay mode, for example, there is the problem that the loss calculation for stop-loss orders is not calculated correctly. For example, if I enter the market and am stopped out at the next 1m bar, Motivewave does not calculate the exit point where the stop was exactly, but the low point of the 1m bar (not even the high point). However, you can narrow down the problem by testing with tick charts. In real trading, the losses are then much higher because you stop out exactly on the tick (+ slight slippage in a highly liquid market).