Best spread to use for backtesting

 

Hey everybody,

Which spread do you recommend when backtesting - current or a particular fixed number? I'm looking for a starting point that I can tweek later depending on the EA, parameter to be optimized (here I usually use profit maximum and maximum drawdown) and currency pair. Thanks! :)

 

I use a spread of 8 (or 0.8 pips) in my backtests

i think it's a realistic number for the current spreads you can get (including commissions)

but this is the absolute minimum
Testing trading strategies on real ticks
Testing trading strategies on real ticks
  • www.mql5.com
The article provides the results of testing a simple trading strategy in three modes: "1 minute OHLC" using only Open, High, Low and Close prices of minute bars; detailed modeling in "Every tick" mode, as well as the most accurate "Every tick based on real ticks" mode applying actual historical data. Comparing the results allows us to assess...
 
mike_lawrence:

Hey everybody,

Which spread do you recommend when backtesting - current or a particular fixed number? I'm looking for a starting point that I can tweek later depending on the EA, parameter to be optimized (here I usually use profit maximum and maximum drawdown) and currency pair. Thanks! :)

I think that it depends on what you want to achieve. If you use a limited value you have the chance that your order will not get filled. But if it gets filled you can be assured of the maximum slippage. On the other hand, if you give it a value as high as possible (ULONG_MAX perhaps) you can be sure that the order will be filled. But the slippage may be larger than you expected.
 
Thanks for the suggestions. I'm kinda a newbie so I just used balance and drawdown as parameters. I started with 8 and then tried 10 and then 15. The results haven't been great. Maybe I need to tweek the other EA inputs more.
 
Jean Francois Le Bas:

I use a spread of 8 (or 0.8 pips) in my backtests

i think it's a realistic number for the current spreads you can get (including commissions)

but this is the absolute minimum

Wow, I thought the value in trat tester was in pips, ie 2=2pips, but you are saying 2=0.2pips?

 
cidand1:

Wow, I thought the value in trat tester was in pips, ie 2=2pips, but you are saying 2=0.2pips?

if you are using 4 digits broker ( I am not sure they still are..) 2 = 2 pips

but in 5 digits broker 2 = 0.2 pips

 
mike_lawrence:

Hey everybody,

Which spread do you recommend when backtesting - current or a particular fixed number? I'm looking for a starting point that I can tweek later depending on the EA, parameter to be optimized (here I usually use profit maximum and maximum drawdown) and currency pair. Thanks! :)

Use a spread a bit wider than you would normally expect.

If your strategy relies on razor thin changes in spread and no slippage to make a profit or not, then you'll need a better strat, so be conservative rather than optimistic in this regards. In this way you can build potential slippage into the spread.

So on EU (eg) id use 3 or 4 full pips (30 - 40 points) of spread.

 
mike_lawrence:

Hey everybody,

Which spread do you recommend when backtesting - current or a particular fixed number? I'm looking for a starting point that I can tweek later depending on the EA, parameter to be optimized (here I usually use profit maximum and maximum drawdown) and currency pair. Thanks! :)

It is better to use current because that gives a true picture of the market condition i.e. how your EA would have performed in real time if you were able to trade during that session/period that was in the past. I demo traded an EA throughout May, then backtested it in June and the result gave 99% accuracy because I used current settings while testing.    

Reason: