Spread and slippage

Dear all,

 

I have a question to be sure my backtesting is accurate enough and, at the same time, is not excesively pessimistic.

 

When we add a market (I trade futures) we indicate spread and slippage, if I want to trade for example ES which usual spread is 1 tick, should I also indicate 1 tick slippage or this slippage caused by the difference between the bid and ask is already considered in the spread? If this is the case then, is the purpose for the slippage field just to retest the system in specially difficult market conditions and should not be used to develop systems?

 

I would like to have this very clear as it really affects the systems in backtest and, most important, during live trading.

 

Thanks.

Hello,

 

in case of ES you could set the slippage to 0 ticks and just use spread of 1 tick. When using market orders you will get in most cases the bid/ask price fills usually. On the other hand when you trade markets like EMD (midCap 400) or TF (Russell 2K) you will get some fill slippage even if you see 1 tick spread most of the time. From my experience these markets are so fast usually that it is very common to have 1 – 3 ticks slippage. When markets get really nasty it can be much bigger value. In this case I would also incorporate some additional slippage i.e. set some slippage value. As nobody is able to provide you with an optimal value the best source to get it is a personal experience.

If you are too strict and set higher value you will have some extra margin for possible worst case scenarios

Hi,

Then, would you consider the following a correct approach?:

– ES market strategy development, spread 1 tick, slippage 0 ticks, then robustness tests using spread 1 tick and robustness 1 tick.

– EMD or TF market strategy development, spread 3 ticks, slippage 0 ticks, then robustness tests using spread 3 ticks and slippage 2 ticks.

Thanks.

Yes, this could be very close to real conditions