Agreed with that, I also use the m30 like you suggested. So you just wait for the colour changing, like the author suggested? and you didn't move the SL at all?
4 HD system (trend follower) 3 replies
Journal for Trend Follower 5 Minute System 9 replies
Trend Follower System Programming 4 replies
A good EA trend follower? 2 replies
trend follower system templates and indicators an issue 5 replies
DislikedSo what do you say Paul are you ready to let us have a go at it? Sounds like its pretty close to right. I think your right though. Surely this week must be some kind of fluke. Those are some great numbers for a trading system! We are drooling over here. Ready to post it on the forum? You're right about the exit strategy. Its pretty much a "to each his own" kind of thing. But thats what The Wicker said at the outset. The system is only part of the success of a trader. The rest is in the psychological makeup of the trader. Like he said, you can have a great system and still not make money with it. So when do we get to play with the new toy? We're all anxious to do some back testing and live trading a demo with it. To bad you werent trading with real money this week! Of course you can fix that next week if you have the nerve.........unleash it....turn that puppy loose! lol............pipip.......Jimdandy.Ignored
DislikedIt just gets better ...
Been waiting all week for it to take a USD/GBP trade - finally just after midnight ilast night t took a short - currently +115 pips up ! So far this week: realised +650 pips unrealised +2220 nearly +3000 for the week as I write - this just cannot be typical - wish I was live with it!!!!!Ignored
Disliked3) Calculate the number of lots to trade (would appreciate input on this) ... I was thinking you could either have the no. lots as an input or say trade n lots per 1000 equity in your account with a minimum of 1 and maximum you define?
4) Make sure it handles where you have two trades open for the same pair in different timeframes - so the right trade get closed if one needs to be closed and the other should remain open
So a few bits outstanding ... but they are not too difficult to do ...
Best
PaulIgnored
extern double Lots = 1.0; extern bool UseMoneyManagement = false; extern bool AccountIsMicro = false; extern int Risk = 0; extern double DecreaseFactor = 0; double CalcolaLot(int Risk) { if(UseMoneyManagement==false) return(Lots); double lot=Lots; int orders=HistoryTotal(); int losses=0; int decimalPlaces=1; if(AccountIsMicro==true) decimalPlaces=2; lot=NormalizeDouble((AccountFreeMargin()*Risk/1000.0)/100,decimalPlaces); if(DecreaseFactor>0) { for(int i=orders-1;i>=0;i--) { if(OrderSelect(i,SELECT_BY_POS,MODE_HISTORY)==false) { Print("Error in history!"); break; } if(OrderSymbol()!=Symbol() || OrderType()>OP_SELL) continue; //---- if(OrderProfit()>0) break; if(OrderProfit()<0) losses++; } if(losses>1) lot=NormalizeDouble(lot-lot*losses/DecreaseFactor,decimalPlaces); } if(lot<0.1 && AccountIsMicro==false) lot=0.1; if(lot<0.01 && AccountIsMicro==true) lot=0.01; if(lot>99) lot=99; return(lot); }
int MagicFromSymbol() { int MagicNumber=0; for (int i=0; i<5; i++) { MagicNumber=MagicNumber*3+StringGetChar(Symbol(),i); } MagicNumber=MagicNumber*3+Period(); return(MagicNumber); }
DislikedIdeas for exit.
Add another HMA(slope direction line) that is a user definable period. Take a look at somewhere around a 15 period to start. Have a trailing stop as well and whichever hits first is out. A little harder but the option to use a traditional MA at a price touch as exit although not as responsive . The option to use either or both would be best.
extern bool usehmaasexit true;
extern bool usetraditionalmaasexit true;
If you had all three one would get you out early enough.
Also a breakeven function in my mind is quite important. I use 1-5 min timeframes mostly and there is NO reason to accept a loss once a trade is up 10 pips or so. I would rather preserve capitol and miss a run than take a loss, and you can always reenter in the direction. Thats the only problem I see with this system so far is getting in on a retrace of a large move but that is another subject.Ignored
DislikedI want to get rid off those blue and red daily, weeky and monthly support/resistance lines. What file do I need to delete?
Thanks,
WimIgnored
DislikedHi Wiper,
As Musketeer suggested, you can get rid of all pivot lines by deleting the Fib Pivots and Pivot Points indicator on your chart, but if you only want to keep lets say daily pivots then go to pivot point properties on the Pivot Point Multi Time Frame click edit and check false on weekly and monthly pivots, now you'll only have daily pivot points on the chart. Hope that helps, have a great weekend!......T - WolfIgnored
DislikedWell all my trades have now closed with a net profit of 1789 pips for the week. For sure this is not going to be typical!!!!!!!!
The one thing I have noticed already is that just using the slope change as an exit works ok when you get a massive move but not so well for faster shorter moves ... e.g. overnightr the EA opened a short USD/GBP position at one point it was 140 points in profit ... in the end it was stopped out for a loss of 34 pips. That's not really acceptable! For the 5min charts The Wicker suggested using the pivots ... looking at the action that is without doubt the way to go (not that I ever doubted his suggestion) ... for the longer term timeframes maybe a trailing stop of some sort might work well ... Billbss (i think) suggested an expotential ma shifted as a trailing stop... that looks pretty interesting - it could certanly close out trades earlier (well depending on settings) than the slope change method - it will likey result in some trades being closed earlier than ideal but looking at it quickly I think it could pull decent profits of of the 30/40/ 50 pip moves that might be more typical than the crazy moves we have had this week.
What do ya all think?
Over the weekend I'll try and finish the trading EA for you to run and have an option to close based on 1) slope change or 2) a MA (you can input the parameters) or 3) an ATR calculated stop .(inputs as desired) ...then maybe someone could pick up the task of backtesting the input options and seeing what look to work best for each pair/timeframe.
PaulIgnored