Each market whether it be fx or indices trading has a spread and this spread changes depending on the time. How do you combat this when designing an automated trading system because depending when your entry is triggered the spread can vary. Fyi I typically trade 1 and 4 hr charts
In reality you can’t. It is a big problem in back testing as spreads back in 1999 will not be the same as they are today – but we only have one setting to use. Spreads at midnight on some markets will be huge and during trading hours very affordable. Some people have written code to try to take this into consideration in their strategies, some try just to open trades only during hours that are low spread hours but the trade might still close during large spread hours.
I tend to back test with a very large spread set as my theory is if the strategy can make profit with a huge spread then it should be even better when the spread is smaller. It also kind of allows for any slippage too. I also have a small piece of code that I put in when back testing that takes off a certain amount of profit for each trade to account for potential overnight fees and then provides an ‘adjusted’ equity graph. It is not 100% accurate but is an extra stress test. If a strategy can make profit with a huge spread and deducting overnight fees too then it stands a fair chance in the real world.
I printed out spread reports every4 hrs and use the worst case in back testing thereby giving me the worst possible scenario. But as you said I would rather over compensate as you dont know when your trade will trigger either in market our out of market hours
I printed out spread reports every4 hrs
Where you seeing these? Please post a link?
Always open to new info sources! 🙂
All I do is print out my watchlist every few hours and I note the range of spreads for each market
The unknown spread situation is also why back testing of small pip scalping strategies is virtually impossible with any level of accuracy.
DAX spread is fixed at 1 pip from 08:00 to 16:30 (UK Time) so I guess if a scalper has a average gain (over many trades in BT with time limit of 8 to 16:30 ) per trade of > 4 pips then it is worth using one of our 25 System limit in Demo to Forward Test.
I would also like to add that if ur working a system on say 4h, 1-3 pips should probably not make or break your strategy…
profit – true as my stop is usually quite large on 4 hr tf and can handle the extra 2 or 3 pip spread. You would be an unlucky bastard if you kept get stopped out and needed 1 or 2 pips more in your stop 🙁
profit – true as my stop is usually quite large on 4 hr tf and can handle the extra 2 or 3 pip spread. You would be an unlucky bastard if you kept get stopped out and needed 1 or 2 pips more in your stop 
Ive experienced that with an optimized stop loss i have several times been 1-2 pips away from stop loss but not been stopped out, its quite fun to watch! just a fun fact lol.. also ofc ive seen price hit my stop just 1-2 pips to far down.. also seen targets been missed/hit by just 1-3 pips. Its fun lol
I coded the MFE trailing stop to include a second trailing stop loss level. So if im just like 3-5 pips away from target my trail would close in tight.
I coded the MFE trailing stop to include a second trailing stop loss level. So if im just like 3-5 pips away from target my trail would close in tight.
How tight? I only ask because tight stops are generally not worth having in the long run. I would have thought you would be better just to say I am very close to my target so I’ll close the position. Moving the stop close just increases your chance of winning less.
I coded the MFE trailing stop to include a second trailing stop loss level. So if im just like 3-5 pips away from target my trail would close in tight.
Is the above amend (to Nicolas original) posted on this site please jebus?
Thank You
GraHal
