Parabolic SAR Expert Advisor | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

4hr Parabolic sar working well 37 replies

Parabolic SAR on MT? 7 replies

Parabolic SAR (Stop and Reverse) ind 7 replies

Parabolic SAR 4 replies

Quick Question about parabolic SAR 1 reply

Parabolic SAR Expert Advisor

  • #1
  • First Post: Dec 17, 2006 6:41pm Dec 17, 2006 6:41pm
Hi guys,

I have been trying to make an expert advisor based on parabolic sar indicators, but been having difficulty in trying to make it work correctly, it will either enter & exit too late or too soon. My idea is to have it action at the very first dot of a parabolic curve.
Can anyone shed some light on this? Or if anyone knows any EA that already uses parabolic sar in it please let me know.

Thanks a lot.
  • #2
  • Dec 17, 2006 6:44pm Dec 17, 2006 6:44pm
  •  tesla
  • Joined Oct 2006 | Status: Friendly Neighborhood Programmer | 533 Posts
Parabolic SAR isn't intended to act as a stand-alone entry indicator. My experience is that it's helpful in identifying exit points but not so helpful with entries.

Lots of people think of things I don't, maybe you'll find a way to make it useful on entries.
  • #3
  • Dec 17, 2006 10:07pm Dec 17, 2006 10:07pm
Quoting tesla
Parabolic SAR isn't intended to act as a stand-alone entry indicator. My experience is that it's helpful in identifying exit points but not so helpful with entries.

Lots of people think of things I don't, maybe you'll find a way to make it useful on entries.
Yes, I agree, I wouldn't use just the parabolic sar to determine entry and exit.
However, my problem is that "how to use parabolic sar in an expert advisor", because according to my experience if I let the EA do automatic trading it keeps taking actions at the wrong time, example:
I instruct the EA to close trade at the first dot of parabolic sar change i.e; parabolic sar changes position from below price to above price and vice versa, but the result is that the EA sometimes exit the trade even without sar changing position or exiting after several dots of position change.

Does anyone have experience of writing an EA using parabolic sar as indicator?
thanks.
  • #4
  • Dec 17, 2006 10:50pm Dec 17, 2006 10:50pm
Since the parabolic change up or down depending on candle high/low vs previous SAR value...
You must not use the current value but the previous value:
- Buffer[1] instead of buffer[0];
The previous value is a fixed one(Like other indicators)

So at a new candle (Open) you check if the previous dot is the first one of the new sequence...

Thats the easiest way to do it...

Ex: Close if candle reach last value(Buffer[1]), enter after first dot(Open of the second candle).
Sorry for my bad english translation
  • #5
  • Dec 18, 2006 1:19am Dec 18, 2006 1:19am
@zauzere

Thanks for the advice, will try to recreate my EA, though am not very sure since I think I have tried this before.

Cheers.
  • #6
  • Edited 12:00pm Dec 18, 2006 5:23am | Edited 12:00pm
Quoting tesla
Parabolic SAR isn't intended to act as a stand-alone entry indicator. My experience is that it's helpful in identifying exit points but not so helpful with entries.

Lots of people think of things I don't, maybe you'll find a way to make it useful on entries.
Really? Did you ask J. Welles Wilder about that because he did described it as a system in his book. SAR stands for "stop and reverse". That certainly sounds like an entry system to me.
  • #7
  • Edited 11:25am Dec 18, 2006 11:02am | Edited 11:25am
  •  tesla
  • Joined Oct 2006 | Status: Friendly Neighborhood Programmer | 533 Posts
Quoting philmcgrew
Really? Did you ask J. Welles Wilder about that because he did describe it as a system in his book. SAR stands for "stop and reverse". That certainly sounds like an entry system to me.
Nope. Haven't asked him.

Like I said in my previous post, lots of people come up with good ideas that I don't. If you think Parabolic SAR can work as a stand-alone entry indicator, I'd love to discuss it.
  • #8
  • Dec 20, 2006 10:51pm Dec 20, 2006 10:51pm
Quoting tesla
Nope. Haven't asked him.

Like I said in my previous post, lots of people come up with good ideas that I don't. If you think Parabolic SAR can work as a stand-alone entry indicator, I'd love to discuss it.
Well, people usually pair the SAR with ADX to eliminate fake entry and exit.
I think Wilder also recommends it in his book.

Update on the EA thing.
Based on Zeuzere explanation, the EA works correctly by using prev period SAR, but entering the market one SAR dot too late often means a losing trade.
Best to stick to SAR in manual trading.
  • #9
  • Dec 20, 2006 11:19pm Dec 20, 2006 11:19pm
Quoting Cancerian
Update on the EA thing.
Based on Zeuzere explanation, the EA works correctly by using prev period SAR, but entering the market one SAR dot too late often means a losing trade.
Best to stick to SAR in manual trading.
Sorry, no, no, no, i did not says that...

I just said, if you use SAR and want to retreive info from it....
use Buffer[1] instead of Buffer[0]...
The way you use SAR is up to you ( I did not talk about EA, just how to use it to have a fix value)...

if you use Buffer[0], you must know how it work(How it was designed:Math stuff) and then react while it happen...
You must know why SAR appear and disappear on the same candle!
_________________________________________________________

Now i'm going to talk about EA ( But its just a example )
If you say after the first dot is to late...
Ex.: Maybe you should place a Buystop (if the trend is down) on the last dot(Buffer[1]):
- lets says you use 15 minutes charts
- You buy or place a buystop on the last SAR value
-- But only if the ADX hourly is Positive(+20) and the -DI is below +DI

But again is just a idea: if the trend on the hourly is up but the short term is down, then you can look to go up, when the SAR turn arround...
Sorry for my bad english translation
  • #10
  • Dec 21, 2006 12:50am Dec 21, 2006 12:50am
ummm... I think I have made a wrong impression here.

Zeuzere i really appreciate your inputs on this, please don't be mistaken, my previous post was never intended as a rebuttal to your post. I am just giving an update on what has happened after I recreate my EA.
Besides, thanks to you, now I know how Parabolic SAR is being calculated by the system.

The Buystop strategy seems interesting, I will look into it furthermore.
Thanks a lot for the advice Zeuzere.

Cheers.
  • #11
  • Dec 21, 2006 1:08am Dec 21, 2006 1:08am
No problemo!
I know that, is just that i want to be clear.
Keep me up to date with you EA, I'm curious

Quoting Cancerian
please don't be mistaken, my previous post was never intended as a rebuttal to your post.
Sorry for my bad english translation
  • #12
  • Feb 3, 2008 9:50pm Feb 3, 2008 9:50pm
any EA here yet?

i have been looking at parabolic sar for weeks now.
i think with a TP, and just using it as entry and exit will work.

you see, manually, and visually, i have found that:

on M15 a TP of just 10, and a double hedge down will result in positive trades (several positions)

on H1 a TP of 50 or so, and double hedge down works. i'm Asian session time, and all the good action takes place while i sleep.
-------

basically, say, 0.1 lot with Parabolic Sar. At hedge point put 0.1 + 0.1. 1st hedge hedges, second one makes back the ground between the two then profit.

if it goes up again, and it does a lot, a little grid appears.
0.1, 0.1, 0.1, 0.1 up at starting point.

vs

0.1, 0.1 down

the TP always gets hit. More trades going in direction of trend.

whenever it leaves TOP zone there are more trades.
whenever it leaves BOTTOM zone there are more trades.


-----top


----- bottom

wherever it goes there are always more trades in that way...
no more trades going against you. watch for overlevering though ig grid gets thick.
  • #13
  • Feb 4, 2008 2:50am Feb 4, 2008 2:50am
@bradman

Can you insert some visual here? I can't follow your explanation. (Sorry)

I haven't touch SAR for trade signals lately, because I found that often SAR gives late signals to entry and exit.

Cheers.
  • #14
  • Mar 14, 2008 7:56pm Mar 14, 2008 7:56pm
Sorry, don't have any screen shots.

But the premise is that there is always more trades going in the direction the currency pair has to go. ie. if it leaves the TOP of your 'grid' you have more open positions going that way, and if it reaches the bottom of your grid, there are more open positions going that way.

At the moment i think i will pay someone to build an EA, and i'm back again to OSMA rather than SAR for entry.

But basically with the grid:

To get here the pair must pass the TOP of your grid

**** HERE




B 0.1
B 0.1
B 0.1
B 0.1
B 0.1



S 0.1
S 0.1
S 0.1


If price goes back to BOTTOM of S then place more Sells (probably 8 gridding down to counter the 5 up Buys.
I calculate the HEDGED loss between the combined B and S and that is countered by having more positions. Equal amount up and down are a simple hedge. Place a few more in the direction the pair is breaking out of your hedge (TOP or Bottom) to break you even from the hedge. And by having more positions you are drawn into profit after a big enough move.

Does get a bit thick if the price only moves a few pips in the day.
That's what EAs are all about.

Of course, it can apply to any grid, not just one launched on SAR or any other indicator.
i prefer a thin grid, so an indicator of good quick entry is best.
But i am going to go OSMA i think
  • #15
  • Mar 28, 2008 1:43pm Mar 28, 2008 1:43pm
This code I use it and it has worked to me, hopefully solves your problem, Greetings. Macbel



Para_0 = iSAR(NULL, 0, 0.02, 0.2, 0);
Para_1 = iSAR(NULL, 0, 0.02, 0.2, 1);


if ((Para_0 <= Ask) && (Para_1<Para_0)) //Signal Buy
{
Signal_B = 1 ;
}
if ((Para_0 >= Bid) && (Para_1>Para_0)) //Signal Sell
{
Signal_S = 1 ;
}
  • #16
  • Mar 29, 2008 7:17pm Mar 29, 2008 7:17pm
parabolic is easy, just use the close

double sar1 = iSAR(NULL, time_frame_for_signal, psar, max, 0);
double cl1 = iClose(NULL, 0, 0);

double sar2 = iSAR(NULL, time_frame_for_signal, psar, max, 1);
double cl2 = iClose(NULL, 0, 1);

if cl1>sar1 itl enter exactly on the dot, and always after

if cl1>sar1 && cl2 <sar2 itl only trade the bar of the exact signal

i made a ea called trendchaser psar
the source is available on the first post of that thread

just search for trendchaser
youl find it
  • #17
  • Mar 29, 2008 7:24pm Mar 29, 2008 7:24pm
my super ea will do exactly that
if you wanna test it
trendchaser super ea
in the trading systems

Quoting bradman
any EA here yet?

i have been looking at parabolic sar for weeks now.
i think with a TP, and just using it as entry and exit will work.

you see, manually, and visually, i have found that:

on M15 a TP of just 10, and a double hedge down will result in positive trades (several positions)

on H1 a TP of 50 or so, and double hedge down works. i'm Asian session time, and all the good action takes place while i sleep.
-------

basically, say, 0.1 lot with Parabolic Sar. At hedge point put 0.1 + 0.1. 1st hedge hedges, second one makes back the ground between the two then profit.

if it goes up again, and it does a lot, a little grid appears.
0.1, 0.1, 0.1, 0.1 up at starting point.

vs

0.1, 0.1 down

the TP always gets hit. More trades going in direction of trend.

whenever it leaves TOP zone there are more trades.
whenever it leaves BOTTOM zone there are more trades.


-----top


----- bottom

wherever it goes there are always more trades in that way...
no more trades going against you. watch for overlevering though ig grid gets thick.
  • #18
  • Apr 10, 2008 9:29pm Apr 10, 2008 9:29pm
Hi
i just pm'ed you.

i'm against hedging now, and use another indicator to filter parabolic sar entry, though still use it for exit.

very unteseted, as i have no ea to backtest, and the indicator obviously can't be backtested.

But i think if i can get the very basic parabolic EA that just buys and sells, then i may be able to code the filter in (or get someone who actually knows what they are doing, to do it).

Links are all deleted in your profile, but i did visit the trencatcher website. i think the new super EA may be a little too advanced with its principles over the simple SAR.
  • #19
  • Apr 11, 2008 5:10pm Apr 11, 2008 5:10pm
I've been trying to develop a PSAR based EA for a few weeks now. Objectives include the ability to stay in the market (or at least "ON") 24/7. The approach here is to set this up then layer items like the ADX to fine tune the entries & exits. Unfortunately, my coding skills aren't up to snuff.

I have set up single direction EA's ... either long or short, but have not yet succeeded in combining them. Anybody want to take a shot at this?

Side comment to Trendchaser. I've just found your thread and am starting to review ... you may like this one. Simple in theory but (I think) very profitable in application. Let me know if you have the time & I'll send the single direction setups to you.
  • #20
  • Apr 12, 2008 9:16pm Apr 12, 2008 9:16pm
i think you didnt search here at ff good
just searching parabolic you woulda found my old thread

this ea can trade psar at its most basic, or advanced

http://www.forexfactory.com/showthread.php?t=54756

Quoting bradman
Hi
i just pm'ed you.

i'm against hedging now, and use another indicator to filter parabolic sar entry, though still use it for exit.

very unteseted, as i have no ea to backtest, and the indicator obviously can't be backtested.

But i think if i can get the very basic parabolic EA that just buys and sells, then i may be able to code the filter in (or get someone who actually knows what they are doing, to do it).

Links are all deleted in your profile, but i did visit the trencatcher website. i think the new super EA may be a little too advanced with its principles over the simple SAR.
0 traders viewing now