Download MetaTrader 5

Backtesting Best Practices

To add comments, please log in or register
Ben Linus
245
Ben Linus  

I've been doing some research online about how to back-test properly and I have a few questions:

  1. I found some sources that explain how to backtest with tick data, without explaining the need for that very much. Is this really necessary for all strategies?
  2. There are other options for the modelling method: "Control points" and "Open prices." When is it OK to use those?
  3. Are there any 3rd party software you would recommend to help backtest better

My strategy is for the daily chart and I wish to test it over the 2004-2014 10 year span. The main logic is applied at the open -- if (Volume[0] <= 1). I didn't even look at lower time frames when developing the strategy.


Advice Welcomed.

Tjipke de Vries
6340
Tjipke de Vries  
BenLinus:


I've been doing some research online about how to back-test properly and I have a few questions:

  1. I found some sources that explain how to backtest with tick data, without explaining the need for that very much. Is this really necessary for all strategies?
  2. There are other options for the modelling method: "Control points" and "Open prices." When is it OK to use those?
  3. Are there any 3rd party software you would recommend to help backtest better

My strategy is for the daily chart and I wish to test it over the 2004-2014 10 year span. The main logic is applied at the open -- if (Volume[0] <= 1). I didn't even look at lower time frames when developing the strategy.


Advice Welcomed.



if (Volume[0] <= 1) will fail with live trading very often check for NewBar

with backtesting you can see what the EA is doing and if it follows the programed strategy

with optimizing you search for a good input combination of your EA

ofcours it is not really necessary for all strategies you can do testing on demo account but to find that way for errors you have found with backtesting you might spend a lot of time

To add comments, please log in or register