What is the difference between the operation models Open Prices Only and the Every Tick?
The operation model indicates how the trading (in live or in backtesting) is performed and what kind of historical data can be used during backtest. The NNFX Algo Tester allows two different operation models, the EVERY TICK and OPEN PRICES ONLY.
EVERY TICK – With this model selected it is possible to open orders 20 minutes before the daily candle closes, according to the NNFX strategy (see TRADING TIME option). You can choose exactly the minute your algorithm should be checked. With this model it is also possible to update the orders (SL TO BE and Trailing SL) outside the trading time (see UPDATE ORDERS ONLY AT TRADING TIME option).
This operation model allows the backtests to be performed accurately, with the disadvantage of slowing down the backtest and is therefore not advisable for parameter optimizations.
In backtesting this model requires 1 min data for the entire backtest period. If you don't have enough historical data then you should use the Open Prices Only model. If you have enough data, the Modelling quality should be equal to or greater than 90%, or n/a but the bar will be completely light green.
OPEN PRICES ONLY – With this model only allows to place trades at the opening of each candle, therefore, place trades 20 minutes before the candle closed is not allowed. In addition, the Trailing SL update and the SL to BE is only done at the candle opening, unlike the Every Tick model. So your algorithm is checked only at 00:00 (TRADING TIME), where orders are placed, SL updated, etc. You cannot change the TRADING TIME.
With this model selected the tests are performed quickly and is the appropriate model for parameter optimization and to use the visual mode. However This model is advisable for beginners.
CONTROL POINTS – This model should not be used as it produces inconsistent results. Please only use Every Tick or Open Prices Only models.
The following tables show the estimated backtesting times with the NNFX Algo Tester. The times depend on the test data, the pair tested but also on the specifications of the computer. It is recommended to use the Every Tick model for single tests and the Open Prices Only model for optimizations.
Note: The backtest times in the table above are estimates only. The test time depends on your algorithm including the indicators used. The use of slow indicators will lead to a longer test time. The test time also depends on the specifications of your computer.