EA based on parabolic SAR... but still unperfect

To add comments, please log in or register
white_tiger
228
white_tiger  

hello, currently i'm developing an EA based on parabolic SAR

the rule is simple,

open a buy position when Parabolic Sar dot lower than current price, hold the posision until the trend reserve (the dot going to higher than current price) & vice versa

i found a big problem for my EA.....

my EA runs very well on the hi-movement price, such as gbpusd last week. But it will become a big problem if the price just ranging (lazy to fly / fall)

i've combined the ParSAR with day_high & day_low to protect my EA for opening position when the chart didn't change too much...

then i put also at my EA, a second ParSAR, you can combine ParSAR step 0,01 with 0,02 for closing a position.

here is information for my EA

step : step for ParSAR #1

max : max for ParSAR #1

step2 : step for ParSAR #2

max2 : max for ParSAR #2

ranging : if day_hi - day_lo > ranging, a position will be opened.

trev : reversal tolerance. if the ParSAR reversed more than (trev) times, the EA will wait for NY market (13.00 GMT) opened or next day Euro market opened (5 GMT) before taking a new position.

trademidnight :

false means, all position will be closed at 01:00 GMT

this EA can modify any position, including position that not opened by it.

result :

standar options :

january 2006 : (1jan-14jan) : 318 pips

you can download the file for back & forward testing until 31 Jan 06.

this EA is for gbpusd 1H

tested at Northfinance MT4

Files:
iwe_ea.ex4 12 kb
sadaloma
709
sadaloma  

Why an EX4 file?

A lot of testers shy away from EX4 for various reasons. IMHO you'll get more feedback if you post the source MQ4 file.

As for myself,no matter how good an EA looks or sound I would not test it if it's in an EX4 format. Furthermore, testing an EA for only 15 days is kind of useless...2 months min before I start to trust it.

It's up to you.

Cheers

PS: BTW if this EA is going to be commercial, you are posting in the wrong section of this forum. the forum's commercial section is at:

https://www.mql5.com/en/forum

Sada

Fuki Nagashi
21
Fuki Nagashi  

I agree with sadaloma. Since MQL has filesystem-capabilities, you can actuall use to plant a virus/trojan or to spy out sensitive information.

No one in his right mind should execute *.ex4 files, which are not self-compiled or out of a REALLY trustworthy source.

Fuki Nagashi
21
Fuki Nagashi  

Oh and ... PSAR sucks for Long/Short Open/Close Signals. They hav some signficance for a dynamic StopLoss, but that is about it.

Really look at one week of trading with PSAR. Yes, the big movements you will get (like with 90% of all the other indicators), but you will burn the money in a market which is not trending like crazy.

safari
26
safari  

Parabolic is indeed a great predictor of turning points

All

I use a custom redesigned parabolic sar i call Kasa to predict turning points. I basically was able to stay clear of most noises, and take signals during active trading hours when the market is likely to make a sustained move. its also great for the indexes

I hope to develope this into EA and this EA could have been a perfect starting point if the owner shared source codes -

See attached graph for kasa turning points

Files:
kasa1.gif 11 kb
white_tiger
228
white_tiger  

hello.....

i just put ex4 last week because the code is very ..... unreadable...

now i have put some part to precedure & function, make us easy to read the source-code. And i've uploaded it.

@safari, how about if we join your custom parsar with my EA

it will be a great

btw, i have deleted the second parsar, because it is useless...

Files:
iwe_ea.mq4 10 kb
Perky
971
Perky  
safari:
All

I use a custom redesigned parabolic sar i call Kasa to predict turning points. I basically was able to stay clear of most noises, and take signals during active trading hours when the market is likely to make a sustained move. its also great for the indexes

I hope to develope this into EA and this EA could have been a perfect starting point if the owner shared source codes -

See attached graph for kasa turning points

can you please post kasa indicator??????

white_tiger
228
white_tiger  

hello,

i have recode my EA, so it will more unreadable..... (i think, it is very clear)

it uses parsar as the main indicator, and only can open a position at 7 GMT -11 GMT and 15 GMT - 19 GMT. the rest, EA can only close a position.

EurUsd M30, please use .set configuration for Expert Properties.

backtest result :

1 january 2006 - 20 january 2006 : 312 pips

dec 05 : 340 pips

please help me to forward-test my EA

welcome to any suggestion for improvement

Files:
ea2.zip 3 kb
sadaloma
709
sadaloma  
white_tiger:
hello,

i have recode my EA, so it will more unreadable..... (i think, it is very clear)

it uses parsar as the main indicator, and only can open a position at 7 GMT -11 GMT and 15 GMT - 19 GMT. the rest, EA can only close a position.

EurUsd M30, please use .set configuration for Expert Properties.

backtest result :

1 january 2006 - 20 january 2006 : 312 pips

dec 05 : 340 pips

please help me to forward-test my EA

welcome to any suggestion for improvement

Thanks for the EA.

I will forward test this EA by tomorrow, will let you know the results the next week.

Sada

white_tiger
228
white_tiger  

please notice that my EA have some bugs...

i gotta fix it soon....

white_tiger
228
white_tiger  

please notice that my EA have some bugs...

i gotta fix it soon....

12
To add comments, please log in or register