fugly's mech system 100 pips a day | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Percentage vs. Pips per Trade vs. Pips per Day/Hour! 7 replies

Simple MACD System - Make 100 Pips a Week! 15 replies

Stop Reverse System 100 pips/day - Expert Needed ! 31 replies

Best 4hr mech system for you? 0 replies

Attachments: fugly's mech system 100 pips a day
Exit Attachments

fugly's mech system 100 pips a day

  • #41
  • Dec 7, 2008 12:38am Dec 7, 2008 12:38am
I have an idea.....

Arlilou,
Is your testing code in mq4 format ? If it is, post it up here for the group and see if anyone can tweek out something
that the rest of us haven't found yet. I actually think you've got it covered fairly well, but, I thought about it and decided
if someone out there could build a better mousetrap, far be it from me to discourage them from doing so.

Even with the experience I've had with this market tendency and tried to share it here in this thread,
it's not impossible that I may have missed something along the way.

Wouldn't it be worthwhile to see what else others come up with ?
If you think the same way, consider contributing the EA to the group and let them tear it apart to look for clues.
If nothing else, it would confirm what we've discussed so far this weekend.
It's worth a look see.

Regards,

Toshi
  • #42
  • Dec 7, 2008 1:35am Dec 7, 2008 1:35am
I do agree with Toshi. I made several tests by hand of the recent history and the results are great.
I would like to see what happened during the last year.
So I hope you will share your EA and let's have a look at it.

I think it is very worthwhile to study this kind of price action.

best regards
Trading S/R with PA only. No indicators anymore!
  • #43
  • Dec 7, 2008 3:31am Dec 7, 2008 3:31am
Take a look at some of the pitfalls in blindly following the concept:

This is Euro from the other day. I ramdomly picked a few bars just to illustrate.

At Point 1, you actually see that for the first 4 minutes the candle was briefly red while below 12625, but no signal yet.
Signal came when the candle turned up breaking the 25 for a positive move of ~75 pips.

At Point 2, the candle fluctuates between up, then down then up again at Point 3 for another positive move.
Depending on how one handled the trade(s), breaking even would have been a welcome outcome. It was a tight call.

At point 4 and 5, same thing. First a fake to the downside at Point 4, red candle at 12725.
Then a pop up to 12750 and the candle is positive by Point 5. It runs almost 100 pips to 12848.

Point 6 is special. The setup for a 3 bar reversal is in place. The candle turns red at 12825.
With no more than 10 pips of drawdown at the 12835 area, the candle pays about 35 pips down to 12790.
Attached Image (click to enlarge)
Click to Enlarge Name: euro_12042008.PNG Size: 41 KB
  • #44
  • Edited 3:58am Dec 7, 2008 3:40am | Edited 3:58am
My point of showing this is that any preprogrammed EA would have likely gotten whipsawed with flat 25 pip stops
at least twice at points 2 and 4 and possibly one other time at Point 3 at 12700 when candle reversed up.

Factor in the 75,100, and 35 pip moves and the profit potential is still there, however, a mechanical EA would likely
get whipped to death at certain times of the session and perform statistically better at other times.
Depending on the broker, spreads could also be a factor, triping a mechanical stoploss order of X number of pips due to spread.
Without good concentration, manual entry/exit could be dangerous because of the human factor:
miss one play and you can wind up missing the ballgame.

A nimble trader could have actually squeezed out pips at points 2 and 4 while still going for the ride at points 3 and 5.
If spreads were whacked out, or if your platform won't allow a limit of <5 pips, it probably wouldn't matter anyway.

So, if there's any statistical edge hidden away in all of this, no doubt it's something to explore.
  • #45
  • Dec 7, 2008 5:45am Dec 7, 2008 5:45am
Ok guys, I made a litle template for this method, and I hope I did it right.

On the chart (wich is 15min tf), you can see a big candels....they are
a 1hr candels so you dont need to switch TFs.
White number in upper right corner is the countdown to the candle close time, and the white text in the left corner (none=0) will display the
difference between the open price of the last bar and the close price of the previous bar. Also I added a grid for the round numbers.

Regards
Attached Image
Attached File(s)
File Type: zip fugly mech system.zip   309 KB | 846 downloads
forexflash
  • #46
  • Dec 7, 2008 6:21am Dec 7, 2008 6:21am
Quoting Arilou
Hi Fugly,

Thanks for sharing this system. I studied your chart of nov.26, but I could find only 56 pips on the chart. I also backtested some weeks in oct. and nov with Trading Simulator. What I found is that the system is probably profitable, but not as much as 100 pips per day. The reason is the spread.

The system gave 40-80 net pips per day, but it required 10-12 trades per day. Considering that one winnig trade costs 3 pips (1 pip+spread) and one losing trade costs 5 pips (3 pips on entry and 2 on stoploss), it is cca 40 pips cost per day....
Hi Arilou,


Thanks for your help, on nov 26 i did get over 200 pips see it depends on what kind of exit strategy you employ, exit half the position at the next horizontal line move stop to breakeven and let it ride, even if you exit everything at the next horizontal line I'm still getting more than 56 pips.

Thanks aslo to your testing, if you can please try on hourlies, it would also be great if you could vary the tp and sl, tp 5 pips + spread and sl 25 pips. Just try it out of curiosity. Even if the stop is larger than the tp it should still work there arer many systems like this check out

http://championship.mql4.com/2008/users/abeiks

I've tested visually of course but i find when one of these levels is breached it tends to follow through.
  • #47
  • Dec 7, 2008 6:24am Dec 7, 2008 6:24am
Quoting FOREXflash
Ok guys, I made a litle template for this method, and I hope I did it right.

On the chart (wich is 15min tf), you can see a big candels....they are
a 1hr candels so you dont need to switch TFs.
White number in upper right corner is the countdown to the candle close time, and the white text in the left corner (none=0) will display the
difference between the open price of the last bar and the close price of the previous bar. Also I added a grid for the round numbers.

Regards
Thanks for your help flash, I hope our members actually forward test this system ....... I'm very optimistic
  • #48
  • Dec 7, 2008 7:49am Dec 7, 2008 7:49am
Quoting FOREXflash
Ok guys, I made a litle template for this method, and I hope I did it right.

On the chart (wich is 15min tf), you can see a big candels....they are
a 1hr candels so you dont need to switch TFs.
White number in upper right corner is the countdown to the candle close time,
and the white text in the left corner (none=0) will display the
difference between the open price of the last bar and the close price of the previous bar.
Also I added a grid for the round numbers.

Regards

Say Flash,

It's late/early so I might not be thinking straight, but wouldn't it be useful for this to display the net change +/- from the previous hour's close?
The difference of the close(1) and the open(0), unless it gaps from a weekend or something would be small.

how about:

double a = ((iClose(NULL,0,0))-(iClose(NULL,0,1)))/Point;

which gives the current price (bid, i think) subtracted from the close of the previous hour.

Sorry if i got it wrong, lemme know.

P.S. countdown timer is a cool idea.
  • #49
  • Dec 7, 2008 7:59am Dec 7, 2008 7:59am
Toshi, here is yours edit......
Attached File(s)
File Type: mq4 Raznost3.mq4   2 KB | 494 downloads
forexflash
  • #50
  • Dec 7, 2008 8:04am Dec 7, 2008 8:04am
Quoting FOREXflash
Toshi, here is yours edit......
Thanks man, got it.

Hey, let's program it to play MP3's
  • #51
  • Dec 7, 2008 8:10am Dec 7, 2008 8:10am
Peace of cake.......
ps. I added some visual efects that looks like light show on your screen!
Attached File(s)
File Type: mq4 Raznostv33 MP3 player.mq4   2 KB | 466 downloads
forexflash
  • #52
  • Dec 7, 2008 8:17am Dec 7, 2008 8:17am
Flash,

How much can we sell this for on Ebay ?

Don't worry Fugly, you'll get a piece of the action, buddy....
  • #53
  • Dec 7, 2008 8:19am Dec 7, 2008 8:19am
Dont know about you, but Im selling it already for 3.95
forexflash
  • #54
  • Dec 7, 2008 8:20am Dec 7, 2008 8:20am
Too late....

I think somebody already is selling it. Shit !
  • #55
  • Dec 7, 2008 8:35am Dec 7, 2008 8:35am
you were right....
Attached Image (click to enlarge)
Click to Enlarge Name: 173918_large.jpg Size: 139 KB
forexflash
  • #56
  • Dec 7, 2008 8:38am Dec 7, 2008 8:38am
Quoting toshi

how about:

double a = ((iClose(NULL,0,0))-(iClose(NULL,0,1)))/Point;

which gives the current price (bid, i think) subtracted from the close of the previous hour.

Sorry if i got it wrong, lemme know.
Ooops, I AM wrong....

iClose(Null,0,1) is the closing price (close) of the previous candle for the default TF of the chart.

iClose(Null,PERIOD_H1,1) is the closing price (close) of the previous 1 HOUR candle, regardless of the chart TF.
  • #57
  • Dec 7, 2008 8:47am Dec 7, 2008 8:47am
here.....
Attached File(s)
File Type: mq4 Raznost2.mq4   2 KB | 476 downloads
forexflash
  • #58
  • Dec 7, 2008 10:41am Dec 7, 2008 10:41am
some tools you could find useful:
Difference(raznost)3 - histoirical current open-prev.close (sep.w)
CC and Cust.CandleShift (builds candles with variable open time) (Mancurt) -very good; one restriction - doesn't support TF>H4
Attached File(s)
File Type: mq4 CustomCandle.mq4   9 KB | 374 downloads
File Type: mq4 CustomCandleCC_Shift.mq4   9 KB | 342 downloads
File Type: mq4 candles_m-Candles_m.mq4   7 KB | 374 downloads
File Type: mq4 Difference_Close1_Open0 _003_sw.mq4   2 KB | 371 downloads
  • #59
  • Dec 7, 2008 12:00pm Dec 7, 2008 12:00pm
Quoting toshi
I have an idea.....

Arlilou,
Is your testing code in mq4 format ? If it is, post it up here for the group and see if anyone can tweek out something
that the rest of us haven't found yet. I actually think you've got it covered fairly well, but, I thought about it and decided
if someone out there could build a better mousetrap, far be it from me to discourage them from doing so.

Even with the experience I've had with this market tendency and tried to share it here in this thread,
it's not impossible that I may have missed something along the...

Hi Toshi,

Unfortunately, I cannot program EA-s, so I wrote the code in Visual Basic.
  • #60
  • Dec 7, 2008 12:26pm Dec 7, 2008 12:26pm
Quoting fugly
Hi Arilou,


Thanks for your help, on nov 26 i did get over 200 pips see it depends on what kind of exit strategy you employ, exit half the position at the next horizontal line move stop to breakeven and let it ride, even if you exit everything at the next horizontal line I'm still getting more than 56 pips.

Thanks aslo to your testing, if you can please try on hourlies, it would also be great if you could vary the tp and sl, tp 5 pips + spread and sl 25 pips. Just try it out of curiosity. Even if the stop is larger than the tp it should still...

Hi fugly,

It is true I tested nov 26 with the simplest exit method, but I think backtesting with trailing stops is almost impossible, because you must go down to tick level data. Remember, trailing stop can move with every tick, so looking at only the 1H chart you simply cannot tell where your trailing stop is and when it gets hit.

Now I examined your nov 26 chart more closely, with tick level data, and I found that the results are even worse with the simple exit rule. Please see:

Candle 1:00: Lose
Candle 3:00: Lose
Candle 4:00: Lose
Candle 5:00: Win
Candle 6:00: Win
Candle 7:00: Win
Candle 8:00: Win
Candle 9:00: Win
Candle 12:00: Lose
Candle 13:00: Win
Candle 14:00: Lose
Candle 15:00: Lose
Candle 17:00: Win
Candle 18:00: Lose
Candle 19:00: Win
Candle 21:00: Win

9 wins and 7 lose. This means 50 pips gain. 9 wins cost 27 pip spread and 7 lose cost 35 pips spread, so the net result is 50-27-35= -12 pips loss, not even 56 pips gain as I first thought. First time I did not use tick level data, only candle fractal data. The devil is always in the details

Now candle 15:00 is tricky, because it seems a winner, but if you watch it closely, you see that it just hits the upper line. And if you test it on tick level data, yes, it happens earlier than hitting the target.

Now I do not know what rules you used, you could reenter this candle later again, and then it was a winner too, but during the test I entered each candle only once.

In the program I tested 1H candles, but broke down the candles to 5M bars to limit the uncertainty. I also experimented with line distances other than 25 pips, but none of them gave better results.

So from the test on one year data it seems that there is absolutely no relation between the signals and the subsequent prie movement.

I know that there are periods, days or even weeks when this method, like every method, gives good results. But seeing the one year profit and loss curve, it is natural. Every P/L curve has rising and falling sections. But what matters is the whole curve itself. And the whole curve points very definitely downwards.
0 traders viewing now