DislikedWhat I meant was if I test a strategy with a fixed TP and SL, I get different TP and SL pips at each order....Ignored
Yes I was testing this morning before I left for work and noticed this. When using OHLC and my profit target is say $100 sometimes it closed at like $400... but also the system doesn't have a hard SL so it would be the same both ways. I think I know what is going on... its because you have multiple symbols being tested at once.
You get 4 ticks from each symbol per tf bar. You will get the Open first then the High then the Low then the Close... since each pair might be in different states in terms of PL at each of the four ticks then it will only close once the entire PL is over your threshold ($100 for example). This only happens at one of the 4 ticks per bar. So if I choose to use 30 minute OHLC for very fast testing it will check 4 times each 30 minutes in the history to see if my profit threshold was met. At the high of my EURUSD long it will also compare that with the High of my GBPUSD short... so it probably won't close because my PL is most likely not $100... but at the Close it will check again and it may have jumped over $100 and to something like $150.
There's nothing stopping you from trading this way... in fact it may be better to use this because if your EA is actually trading with the trend then chances are you will get into a situation where you are in profit bigtime instead of taking your small 1% cut you will probably get more. Ofcourse this comes at the expense of those situations where you barely get to your 1% only to reverse and go into DD where each tick test would closed in profit.
I found testing this way increases my profit by a lot because it gets you more than x% profit each time... while maintaining similar DD levels as the tick test(this is the key)... so basically what this is saying is that we aren't taking enough profits we are happy with small profits at the expense of a large potential DD.
I'm now testing 7% PP to see if I can get through 2010 with the same DD as using 1% PP.
If you change the EA to look at closed bars only this may help profitability but also increase chance for DD... if the backtest actually survives then this may be the better option(aggressive setting)... instead of using a hard profit percentage... get out when you are in profit on a closed bar.
So testing this way actually ensures stability more than anything else... since we always trade at open of each bar the entries are all the same.. but out exits are more loose with OHLC or Mt4 closed bars. If we get a margin call then chances are the EA will get you a margin call with each tick aswell!
Yes 0.001 lots are def a plus .. it used to be min 1 standard lot... for testing doesn't matter since we can increase our starting equity... but good to know if you are going live with this broker (I still don't recommend it).
The beauty of this platform is it was both OnBar() and OnTick() so however you program your strategy it will act the same regardless of the type of backtest you do... unlike MT4.
I added all my new trade logic in OnBar and check for profit in OnTick.
Jag