- Home
- Community
- Handelssysteme
- GCG - 'J36' Auto Scalper Test
GCG - 'J36' Auto Scalper Test (bei jimloko)
Gewinn : | +58.28M% |
Drawdown | 7.53% |
Pips: | 474917.0 |
Trades | 428992 |
Gewonnen: |
|
Verloren: |
|
Typ: | Demo |
Hebel: | 1:1000 |
Trading: | Automatisiert |
Edit Your Comment
GCG - 'J36' Auto Scalper Test Diskussion
Mitglied seit Oct 03, 2013
2 Posts
May 24, 2014 at 16:36
Mitglied seit Oct 03, 2013
2 Posts
i am interested in dis EA, pls send me mail at [email protected]. thnks
Mitglied seit May 02, 2014
5 Posts
Mitglied seit May 02, 2014
5 Posts
May 26, 2014 at 20:47
Mitglied seit May 02, 2014
5 Posts
I noticed this is 1M strategy where you sell/buy closer to the high/low of the bar and close at 1pp take profit. How do you define the direction of the 1st trade in each minute, because there isn’t much history in that minute?
Any intel is highly appreciated.
Any intel is highly appreciated.
May 27, 2014 at 12:21
Mitglied seit Nov 30, 2013
23 Posts
Hello,
This EA only works on Demo with the broker GCG, which aparently is a bucket shop!
@ltrading2007 I basically don't worry about the candle and only use the tick data. Basically i find the mean of the last n ticks. Then if price ticks below this I buy and expect price to move back to the mean, or centre of the candle, then typically it ticks to the other end, i then close with a 1pip profit.
cheers
This EA only works on Demo with the broker GCG, which aparently is a bucket shop!
@ltrading2007 I basically don't worry about the candle and only use the tick data. Basically i find the mean of the last n ticks. Then if price ticks below this I buy and expect price to move back to the mean, or centre of the candle, then typically it ticks to the other end, i then close with a 1pip profit.
cheers
Mitglied seit May 02, 2014
5 Posts
May 27, 2014 at 13:43
Mitglied seit May 02, 2014
5 Posts
Thanks for sharing.
I noticed GCG has 10k accounts with 0.5 spread. With spreads that low there might be an application here assuming lots get capped at 10-50, we'll have to try.
Is there an approach you are using when defining the N variable or is it more or less stable? I imagine N is not that long; otherwise it relies too much on mean reversion of the price.
Also, do you wait for the price to get to a certain distance from the mean? Here Bollingers with shorter N might be of good use.
The stable mean reversion in that particular time frame is the key to great success in our friend’s video. (
I’ve done a great deal of mean reverting /pair trading in stocks, using from simple Bollingers to hardcore co-integration stats (ADFs). I was looking to apply them in FX, however I’m not sure how well MT4 will handle all this quant luggage.
Cheers,
I noticed GCG has 10k accounts with 0.5 spread. With spreads that low there might be an application here assuming lots get capped at 10-50, we'll have to try.
Is there an approach you are using when defining the N variable or is it more or less stable? I imagine N is not that long; otherwise it relies too much on mean reversion of the price.
Also, do you wait for the price to get to a certain distance from the mean? Here Bollingers with shorter N might be of good use.
The stable mean reversion in that particular time frame is the key to great success in our friend’s video. (
I’ve done a great deal of mean reverting /pair trading in stocks, using from simple Bollingers to hardcore co-integration stats (ADFs). I was looking to apply them in FX, however I’m not sure how well MT4 will handle all this quant luggage.
Cheers,
May 28, 2014 at 08:33
Mitglied seit Nov 30, 2013
23 Posts
ltrading2007 posted:
Thanks for sharing.
No Probs. Capping the lots wouldn't be a problem.
As for N i was planning on making it dynamic based on the number of ticks received over a set period so during slow times it would increase etc. However, I've found that a fixed value works fine, i'm currently using a tick count of 15. And it is basically mean reversion, but on a much much smaller scale, intra candle. So the mean basically moves with the candle so its pretty stable this way. I've noticed that on average bars are at least 3 pips high, so i buy when it pops to the low side and sell when in profit - which is when it pops up to the high side ~2pips resulting in 1 pip profit.
I was planning on using an average to calculate mean distance, but it works so well as is i just keep a dynamic upper and lower limit based on the last 20 candles, and if any open trades full outside of this i just close it - i.e. if a buy has been opened at the top of a retrace for eg and price move away enough i just close. By the time this happens i've completed many trades so the loss on this (or these) is covered. So i don't check distance from the mean: well for this current EA! I am looking some sort of mean reversion for the next iteration which hopefully I can use on other brokers including 5 digit. Co-integration actually sounds rather like an idea I have thought of but wasn't sure how to implement, so i think i'll be reading up!! How has your performance been using MR in the past?
MT4 isn't the quickest platform and will slow down quite a lot with polynomial regression on a few open charts. However, there's no reason why another system can't crunch the numbers and then use MT4 to place the trades!
I have found that I do the best manual trading with breakouts, and I like trading Bollinger breakouts - I find this quite effective. I also like BB during consolidation when combined with PA. I'll probably work up an EA at some point but i have been intrigued by Bader's scalping method and like the idea of many small wins instead of going for less but larger wins.
Ok i'm waffling a bit now :)
cheers
May 28, 2014 at 08:34
Mitglied seit Oct 23, 2012
355 Posts
its not hard to figure it out on those type of charts... 😝
https://www.myfxbook.com/members/sirius1fx/paj-high-frequency-poc/931085
https://www.myfxbook.com/members/sirius1fx/paj-high-frequency-poc/931085
if you follow the flock like sheep you always end up stepping in shit!
Mitglied seit May 02, 2014
5 Posts
*Kommerzielle Nutzung und Spam werden nicht toleriert und können zur Kündigung des Kontos führen.
Tipp: Wenn Sie ein Bild/eine Youtube-Url posten, wird diese automatisch in Ihren Beitrag eingebettet!
Tipp: Tippen Sie das @-Zeichen ein, um einen an dieser Diskussion teilnehmenden Benutzernamen automatisch zu vervollständigen.