Welcome to Forex-TSD!, one of the largest Forex forums worldwide, where you will be able to find the most complete and reliable Forex information imaginable.
From the list below, select the forum that you want to visit and register to post, as many times you want. It’s absolutely free. Click here for registering on Forex-TSD.
Exclusive Forum
The Exclusive Forum is the only paid section. Once you subscribe, you will get free access to real cutting-edge Trading Systems (automated and not), Indicators, Signals, Articles, etc., that will help and guide you, in ways that you could only imagine, with your Forex trading.
Elite Section
Get access to private discussions, specialized support, indicators and trading systems reported every week.
Advanced Elite Section
For professional traders, trading system developers and any other member who may need to use and/or convert, the most cutting-edge exclusive indicators and trading systems for MT4 and MT5.
There is one problem with your test,if I have understood it right,and the problem is that just exiting at opposite wpr is not a good thing as I told you in my previous post...
Can you post the results,for all the entries,and exit just at change in Chand1h or satl4h ?
PLEASE SEE ATTACHED PICs..the pics are in 5m because visually I couldn`t put the complete sequence in 1m,just to try to convey what I mean..
first pic you have 2 entries,the max distance from first(and ,in this case second too) entry was approx 105 pips,and the exit trough chandelier was with approx 50 pips profit...
second pic,1 entry max dist 134 pips,exit by chand 60 pips..
This way we will know more about the average percourse of price and at what point it can be useful to "design" a tp strategy based or not on chand,ie,on volatility...for 5m and for 1m tfs
If you can without doing any extra effort,just post,if not,don`t worry,I will do it during next week..Hopefully..
There is one problem with your test,if I have understood it right,and the problem is that just exiting at opposite wpr is not a good thing as I told you in my previous post...
Can you post the results,for all the entries,and exit just at change in Chand1h or satl4h ?
PLEASE SEE ATTACHED PICs..the pics are in 5m because visually I couldn`t put the complete sequence in 1m,just to try to convey what I mean..
first pic you have 2 entries,the max distance from first(and ,in this case second too) entry was approx 105 pips,and the exit trough chandelier was with approx 50 pips profit...
second pic,1 entry max dist 134 pips,exit by chand 60 pips..
This way we will know more about the average percourse of price and at what point it can be useful to "design" a tp strategy based or not on chand,ie,on volatility...for 5m and for 1m tfs
If you can without doing any extra effort,just post,if not,don`t worry,I will do it during next week..Hopefully..
Thanks again and regards
Simba
I understand what hopefully means Here it is, it didn't even hurt... Exits are always as soon as Chandelier or SATL change their color (it is almost always Chandelier). The trades that would have had to be carried over a weekend are marked. Beware that I didn't modify entries so there are no entries during nighttime, no entries Friday afternoon and no entries just before/after news.
Still it should serve as a good overview of how the system reacts to trends. Would have made some real nice profits, although I'd like to see someone with such guts to trade it live
Looking forward to new comments and conlusions, till then bye
__________________
Malcik
Last edited by Malcik; 02-18-2007 at 06:07 PM.
Reason: Chand table updated in the backtest
Thanks a lot for your effort..the results are impressive..basically I will sum them up,I understand that 3months,150 trades is not enough,but until we have the EA,if we have it ,we(until now we=Malcik..so thanks again) have to test manually..and it takes a lot of time..so,for now I would say that the 1m version is promising and the first candidate for EA testing on 1m ,2 and a half years of data.
2-Profit Factor: 2.95...very good,in 3 months impressive..we have to check on 2.5 years basis
3-Max drawdown PEAK TO BOTTOM in %..we have 2 very similar runs -19.8% and -20.6%...it is high..we have to work on it,by either reducing the max accumulated positions(free until now ,even for losing trades until stopped the system kept accumulating lots on each signal..which in itself means that the chandelier keeps you safe even with excessively leveraged positions),or including some kind of filter for additional positions ie:do not open new one unless at least x pips distance from the first
4-%W= 36%..%L= 64%..the percentage winning trades is low-probably point 3 above modifications will help in improving it-and some form of Tp,even partial must be devised to improve it,the good results come from riding the trend..to the hilt..net Average(for all trades,positive+negative) per trade (25,67 pips),Average W and Average L are,in my opinion non pertinent since there is too much deviation to extract meaning from an average
5-Basically the system,and excuse the expression,runs by runs..from 2000usd to 1604(it starts overleveraging and then Chand cuts it off),then from 1604 to 7421(it starts earning money and overleveraging until stopped out by Chand at a big profit),then from 7421 to 5890..This is extremely interesting,once we have the full test for 2.5 years of data on 1m,we can try to devise some additional filter to improve the results
There is additional interesting comments that Malcik wants to post,and I think these will help us understand better what can we expect and how to modify our system
Thanks a lot for your effort..the results are impressive..basically I will sum them up,I understand that 3months,150 trades is not enough,but until we have the EA,if we have it ,we(until now we=Malcik..so thanks again) have to test manually..and it takes a lot of time..so,for now I would say that the 1m version is promising and the first candidate for EA testing on 1m ,2 and a half years of data.
2-Profit Factor: 2.95...very good,in 3 months impressive..we have to check on 2.5 years basis
3-Max drawdown PEAK TO BOTTOM in %..we have 2 very similar runs -19.8% and -20.6%...it is high..we have to work on it,by either reducing the max accumulated positions(free until now ,even for losing trades until stopped the system kept accumulating lots on each signal..which in itself means that the chandelier keeps you safe even with excessively leveraged positions),or including some kind of filter for additional positions ie:do not open new one unless at least x pips distance from the first
4-%W= 36%..%L= 64%..the percentage winning trades is low-probably point 3 above modifications will help in improving it-and some form of Tp,even partial must be devised to improve it,the good results come from riding the trend..to the hilt..net Average(for all trades,positive+negative) per trade (25,67 pips),Average W and Average L are,in my opinion non pertinent since there is too much deviation to extract meaning from an average
5-Basically the system,and excuse the expression,runs by runs..from 2000usd to 1604(it starts overleveraging and then Chand cuts it off),then from 1604 to 7421(it starts earning money and overleveraging until stopped out by Chand at a big profit),then from 7421 to 5890..This is extremely interesting,once we have the full test for 2.5 years of data on 1m,we can try to devise some additional filter to improve the results
There is additional interesting comments that Malcik wants to post,and I think these will help us understand better what can we expect and how to modify our system
Simba
Here's another update... at the end of the file you can see two tables with what profit the system makes when only allowing for entries with price within 10-20, 20-30 etc pips ahead of the Chandelier. It also shows whether or not it is sensible to take entries below/above SATL
Here's another version, I have added what would happen if we combined filters by distance to SATL and to Chandelier.
With Simba we agreed that we are close to some conclusions that can be passed to the level of 2,5 years backtest and if it confirms, we can make them official
We only allow for long entries when price is above Chandelier by 10 to 50 pips
We only allow for long entries when price is above SATL (by any number of pips).
If you are new here, for other rules please do your least to read through this thread, it is worth it
For short positions we use the exact contrary. Please check the file - the related calculations are to the back of the list.
I coded an EA for M1 SimbaSystem with the addition I mentioned in my last post. I'd like everybody who is willing to contribute to try to test it so that we can get rid of any possible flaws. When this is done we can use it for preliminary testing on 2,5 years 90% quality data. It only has SATL, HULL MA, and JMA in it. You can select which one is to be used (MA_Type 1, 2, and 3 in the order I mentioned the MAs). For HULL MA and JMA you can enter period and for JMA also the Phase parameter. If the particular MA doesn't need period or phase, it will be ignored.
All other parameters are hard-coded for now, except the time of the news which remained there from the previous version I had for myself - just leave it at 1970 and it won't do anything. TP is hard-coded at 30 pips. For reentries, see my last post. Other rules are basic Simba's, except you get to select MA. When we get rid of entry flaws (shouldn't take too long, the EA is simple), I'll make more parameters changable and, if necessary, make other changes.
This won't be as professional and flexible as multixpMAEA (e.g. anytime we want to use something else than SATL, HULL, and JMA, or some other rules, it needs to be updated), but it should be better than nothing for now.
So please, if you can spare 30 minutes for the sake of this great system and for Simba who spends time posting about it, please run it on a few weeks' data and visually check if all entries are correct and there are no missed ones. Post your insights so that together we can debug this EA as soon as we can and start testing.
Next, if you have 2,5 years of 90% quality data, please contact us. EA backtesting is not a thing that takes many hours like manual testing and it would help us a lot. Unfortunately, I'm unable to do it, because I just can't get my MT4 to remember my historical data before September 2006 for time frames M5 through H1. H1 is used by Chandelier, so it jams my quality. The funny thing is that M1 and H4 through MN go all the way back where they should, it's just M5 through H1 that don't. If someone knows the solution, please help
Last thing: please feel free to study the code, it's no secret and I'm not a professional programmer. But please, don't change it, only make proposals, tell me what's wrong and how you would change it. I'll do it, I don't want us again to get confused by several different versions as it is usual. If there's someone who is a better programmer and would like to take care of this job, I'll be glad to leave my job up to him
I coded an EA for M1 SimbaSystem with the addition I mentioned in my last post. I'd like everybody who is willing to contribute to try to test it so that we can get rid of any possible flaws. When this is done we can use it for preliminary testing on 2,5 years 90% quality data. It only has SATL, HULL MA, and JMA in it. You can select which one is to be used (MA_Type 1, 2, and 3 in the order I mentioned the MAs). For HULL MA and JMA you can enter period and for JMA also the Phase parameter. If the particular MA doesn't need period or phase, it will be ignored.
All other parameters are hard-coded for now, except the time of the news which remained there from the previous version I had for myself - just leave it at 1970 and it won't do anything. TP is hard-coded at 30 pips. For reentries, see my last post. Other rules are basic Simba's, except you get to select MA. When we get rid of entry flaws (shouldn't take too long, the EA is simple), I'll make more parameters changable and, if necessary, make other changes.
This won't be as professional and flexible as multixpMAEA (e.g. anytime we want to use something else than SATL, HULL, and JMA, or some other rules, it needs to be updated), but it should be better than nothing for now.
So please, if you can spare 30 minutes for the sake of this great system and for Simba who spends time posting about it, please run it on a few weeks' data and visually check if all entries are correct and there are no missed ones. Post your insights so that together we can debug this EA as soon as we can and start testing.
Next, if you have 2,5 years of 90% quality data, please contact us. EA backtesting is not a thing that takes many hours like manual testing and it would help us a lot. Unfortunately, I'm unable to do it, because I just can't get my MT4 to remember my historical data before September 2006 for time frames M5 through H1. H1 is used by Chandelier, so it jams my quality. The funny thing is that M1 and H4 through MN go all the way back where they should, it's just M5 through H1 that don't. If someone knows the solution, please help
Last thing: please feel free to study the code, it's no secret and I'm not a professional programmer. But please, don't change it, only make proposals, tell me what's wrong and how you would change it. I'll do it, I don't want us again to get confused by several different versions as it is usual. If there's someone who is a better programmer and would like to take care of this job, I'll be glad to leave my job up to him
Bye for now
Hi Malcik
Congratulations again for your exceptional dedication
I will dowload and check the ea for bugs
Thanks,and thanks for the explanations..
Exits are always as soon as Chandelier or SATL change their color (it is almost always Chandelier). The trades that would have had to be carried over a weekend are marked. Beware that I didn't modify entries so there are no entries during nighttime, no entries Friday afternoon and no entries just before/after news.
Linear Mode
