greed999,
Since you feel you are missing entries I would suggest trying to debug the code to understand why it decided not to trade when you thought it should trade. I suggest you acquire a Market Replay recording for the date by going into "File>Utilities>Download Replay Data" and then you can use it to keep playing the same data over and over so you can analyze what exactly is happening and gain insights into why it is or is not trading.
-
Going from backtest to simulation
Here's the situation
At first there was an idea... a mathmetical and logical sequence... on paper it works fine... then there was some coding, backtesting and optimization... several weeks past and it seems to be working... now we are going live-simulated before going live... and strategy seems to be having a day off... instead of 1000 orders per day there are hardly 20... giant slippage... missing in and out conditions... and etc...
the question is, are there any basic rules for fine-tunning, checking and so on? seems like we're missing some huge part of control and analytics...
I do understand that there's a difference between backtesting and simulation (have written couple of backtesters myself and know the way it works from inside) but what to look for?
any ideas? thanks in advance
Leave a comment: