My Live Trades | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Dukascopy Live Account - Don't Post if you haven't traded live with them! 117 replies

My Live Trades on my Live Account 588 replies

go live or bust ... diary of an fx live-trader-wannabe 1,621 replies

Trade 2 live accounts - one with full balance, 2nd as a live test? 0 replies

Oanda 10$ live acc quick-start guide (for newbie to trade live safe) 13 replies

Attachments: My Live Trades
Exit Attachments

My Live Trades

  • #101
  • May 20, 2007 2:15pm May 20, 2007 2:15pm
the elements you see on the charts are:
1. a linear regression, such as y = a + bx, with +/- 2 standard deviations (pink line and green/yellow dotted lines) and
2. a non-linear regression, in this case parabolic, such as y = a + bx + cx(2); that will be the blue line and its +/- 2 standard deviations as continuous green/yellow lines. i also tried a cubic regression, but it is too sensitive to price change.
3. sometimes you will see vegas tunnel indicator, i'm using it to better identify entries/exits.
my general rules: i set my entries when price hits a green/yellow line and i don't usually go against the trend. that's pretty much it for now.
i will post some other charts as soon as i have time to do it.

Quoting emda
yes.. I was going to tell you the same.... you put 3 charts... but don't tell us about the 1 week data chart.

I'll be watching your charts . It's going to be interesting to see how they change over the time. I'm very interested in non linear regressions too....

so... you only set the trades when both linear (pink) and parabolic(cyan) agree and tell you the same? Or you put the pink only as a reference to show us the difference between linear and parabolic?

I'm curious about the red dot...

congratulations for your past trades and this journal.
Nihil Sine Deo.
  • #102
  • Edited 2:51pm May 20, 2007 2:38pm | Edited 2:51pm
1. Linear regression (pink) line shows the main trend is up (3 months data analysis), and that price has already hit a low; now it should return to the pink line: 2.0049. Let's look at the pink line on the 4 weeks chart: it's down, and the price is very close to it. This means it can go either ways, but I think longs are somehow favored for now. My entries:
- SELL @ 1.9839/60, TP @ 1.9680/65
- BUY @ 1.9670/65, TP @ 1.9820
The short term trend is down, although the long term trend is up. Look at both charts.

2. Parabolic regression line (blue) shows the main trend as pointing down on both charts. Same entries as above.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: g-u_4wk_2stdev.jpg Size: 80 KB
Click to Enlarge Name: g-u_3mo_2stdev.jpg Size: 59 KB
Nihil Sine Deo.
  • #103
  • May 20, 2007 3:38pm May 20, 2007 3:38pm
Same thing. Look at both charts, longs are favored short term.
- BUY @ 0.8207/0.8189, TP @ 0.8325
- SELL @ 0.8348/59, TP @ 0.8216
Attached Image(s) (click to enlarge)
Click to Enlarge Name: a-u_4wk_2stdev.jpg Size: 77 KB
Click to Enlarge Name: a-u_3mo_2stdev.jpg Size: 61 KB
Nihil Sine Deo.
  • #104
  • May 20, 2007 3:45pm May 20, 2007 3:45pm
Hey Scrat

It would be clearer to have different colors for the std.dev lines in each algorithm. I almost died looking at your last chart

Here's my sample AUDUSD chart. Red stuff is vegas, linear reg is the thick blue line and thick grey lines are the stdevs derived from it, dim grey for the SMA bollinger.

As you can see I am long now, and should have been 40 pips ago if I had my head screwed on at the time

Regards and sorry for stealing journal space, but it is for the public good
Attached Image (click to enlarge)
Click to Enlarge Name: shot.jpg Size: 80 KB
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #105
  • May 20, 2007 3:59pm May 20, 2007 3:59pm
i think all /usd pairs go long: e-u, g-u, a-u, n-u. the price just bumped into the lower green line (see 3 months data analysis; it's similar for all /usd pairs).
u-chf goes short.
g-j should go long... hm.
let's see if this works.

Quoting Plutonite
Hey Scrat

It would be clearer to have different colors for the std.dev lines in each algorithm. I almost died looking at your last chart

Here's my sample AUDUSD chart. Red stuff is vegas, linear reg is the thick blue line and thick grey lines are the stdevs derived from it, dim grey for the SMA bollinger.

As you can see I am long now, and should have been 40 pips ago if I had my head screwed on at the time

Regards and sorry for stealing journal space, but it is for the public good
Nihil Sine Deo.
  • #106
  • May 20, 2007 4:16pm May 20, 2007 4:16pm
Quoting Scrat
i think all /usd pairs go long: e-u, g-u, a-u, n-u. the price just bumped into the lower green line (see 3 months data analysis; it's similar for all /usd pairs).
u-chf goes short.
g-j should go long... hm.
let's see if this works.
Yep, I am long with you on all those, GBPJPY and EURJPY also, taking profits around 239.80 and 164.15 respectively. G-U I am only long for about 30 pips though, but n-u and a-u are heavy trades.

Good week to you all!
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #107
  • May 21, 2007 6:33am May 21, 2007 6:33am
wrong decision with my longs.
i went long again on e-u, it hit the lower green line. see in the attached photo, my entry is at the horizontal blue line, and my target is at the curved light blue line between the arrows (because this looks like a trade against the trend).
target: probably 1.3515 by the end of the week, but it depends on how e-u is moving
Attached Image (click to enlarge)
Click to Enlarge Name: e-u.jpg Size: 54 KB
Nihil Sine Deo.
  • #108
  • May 21, 2007 6:42am May 21, 2007 6:42am
Yeah, I had noted it was in a transitional chance in direction. So I had keep my shorts open. Woke up in pretty good profit this morning.
I think its going futher in the long term, so I 'm favoring the shorts on the bump and grind down


Quoting Scrat
wrong decision with my longs.
i went long again on e-u, it hit the lower green line. see in the attached photo, my entry is at the horizontal blue line, and my target is at the curved light blue line between the arrows (because this looks like a trade against the trend).
target: probably 1.3515 by the end of the week, but it depends on how e-u is moving
  • #109
  • May 21, 2007 6:48am May 21, 2007 6:48am
and another short on u-chf. my entry is at the horizontal blue line, it'll touch the curved light blue line by the end of week, that is my target since this is a trade against the trend.
Attached Image (click to enlarge)
Click to Enlarge Name: u-chf.jpg Size: 59 KB
Nihil Sine Deo.
  • #110
  • May 21, 2007 7:05am May 21, 2007 7:05am
You've got guts kid. Good Luck! Hope it works out for all of us
  • #111
  • May 21, 2007 7:09am May 21, 2007 7:09am
Lots of SL hit today for me too unfortunately, this is what we get for anti-trend trading, but thankfully I had a sell stop triggered for 60 pips now on e-u. Hope you managed to make some on this unheavenly day scrat.

I am still long GJ, but short EJ and NZDUSD. The kiwi trade sucked big time, but again, our fault for going long against the current without news or psychological/fibo support(i.e vegas). We could have TP at 25 pips (regression mean), but the human is a greedy animal. Silly mistake, never again.

Regards,
-P
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #112
  • May 21, 2007 7:41am May 21, 2007 7:41am
i should have looked at the 3 months data analysis. i found that the regression midline seems to be hit once a week (consistently), but the trades look a little bit different than what i see on my usual chart.
i'll give it a try and see if i can find a way to filter my trades.
here is what i see:
1. e-u long (only now, not 12 hours ago), trend is down;
2. u-chf short, trend is up.
3. i will short g-chf @ 2.4350 area
4. i will long g-u @ 1.9620 area for at least 50 pips.
what not to trade right now:
1. g-j
2. e-j
3. u-cad (although it might be a good idea to short it, i'm not comfortable trading this pair anymore).

Quoting Plutonite
Lots of SL hit today for me too unfortunately, this is what we get for anti-trend trading, but thankfully I had a sell stop triggered for 60 pips now on e-u. Hope you managed to make some on this unheavenly day scrat.

I am still long GJ, but short EJ and NZDUSD. The kiwi trade sucked big time, but again, our fault for going long against the current without news or psychological/fibo support(i.e vegas). We could have TP at 25 pips (regression mean), but the human is a greedy animal. Silly mistake, never again.

Regards,
-P
Nihil Sine Deo.
  • #113
  • May 21, 2007 8:16am May 21, 2007 8:16am
are the deviations always calculated fixed ? Maybe to improve it they could adapt to the weekly range average or something.... just an idea...
  • #114
  • May 21, 2007 8:30am May 21, 2007 8:30am
no, they're changing all the time, because the regression curve is changing. it's a function of closing price values.

Quoting emda
are the deviations always calculated fixed ? Maybe to improve it they could adapt to the weekly range average or something.... just an idea...
Nihil Sine Deo.
  • #115
  • May 21, 2007 9:24am May 21, 2007 9:24am
Quoting Scrat
i should have looked at the 3 months data analysis. i found that the regression midline seems to be hit once a week (consistently), but the trades look a little bit different than what i see on my usual chart.
i'll give it a try and see if i can find a way to filter my trades.
here is what i see:
1. e-u long (only now, not 12 hours ago), trend is down;
2. u-chf short, trend is up.
3. i will short g-chf @ 2.4350 area
4. i will long g-u @ 1.9620 area for at least 50 pips.
what not to trade right now:
1. g-j
2. e-j
3. u-cad (although it might be a good idea to short it, i'm not comfortable trading this pair anymore).
Thx for that tip. I think you are right about not trading the crosses, and I actually exited the e-j trade upon reconsideration. I will give GJ some time to behave, considering I am long with the trend and we might get help from GBPUSD soon as you noted.

I am not trading USDCAD either. Very strong downtrend. Just a sinful little long for 40 pips, it's not much to ask I swear

Good luck.
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #116
  • May 21, 2007 10:08am May 21, 2007 10:08am
Scrat, Plutonite,

Great thread, I have just joined FF and have been following this through.
I've just opened a demo acct with North Finance so that I can get Linear and Parabolic regression lines on charts.
Two questions if I may,

1) How do I know when and where to place my linear and Parabolic regression lines, is it every 4hrs draw a new one? Or is it from the latest extreme price ie: hi or lo?
2) I can't get Parabolic regressions on my chart - can't find the indicator anywhere?

OK three questions....
3) Scrat, have you backtested this way of trading over the last 7 - 10 yrs to see how it performed - without changing any of the parameters as you went along? If so, how did it perform? (I think I've snuck in another one there).

Cheers.....
  • #117
  • May 21, 2007 11:21am May 21, 2007 11:21am
you can't find this indicator in MT's list and i prefer not to disclose it at this moment. as for the backtest, it's not relevant because this indicator changes with the new data. if you backtest, you'll get a perfect record and this is not the case here. i tried this before i started to use it, and i've got 5000 pips for 2 weeks, working on 6-7 pairs if i remember well. too good to be true.
i've been using it for 3 weeks now and my net profit is something between 15-20% up to date. but i still don't know how to use it.
as soon as i come with a winning strategy, i'll post it. it's just that it's gonna take more than i thought.
remember, this thread is a journal and i'll discuss my trades mostly, not my method/system.

Quoting Westozzie
Scrat, Plutonite,

Great thread, I have just joined FF and have been following this through.
I've just opened a demo acct with North Finance so that I can get Linear and Parabolic regression lines on charts.
Two questions if I may,

1) How do I know when and where to place my linear and Parabolic regression lines, is it every 4hrs draw a new one? Or is it from the latest extreme price ie: hi or lo?
2) I can't get Parabolic regressions on my chart - can't find the indicator anywhere?

OK three questions....
3) Scrat, have you backtested this way of trading over the last 7 - 10 yrs to see how it performed - without changing any of the parameters as you went along? If so, how did it perform? (I think I've snuck in another one there).

Cheers.....
Nihil Sine Deo.
  • #118
  • May 21, 2007 12:04pm May 21, 2007 12:04pm
Quoting Scrat
you can't find this indicator in MT's list and i prefer not to disclose it at this moment. as for the backtest, it's not relevant because this indicator changes with the new data. if you backtest, you'll get a perfect record and this is not the case here. i tried this before i started to use it, and i've got 5000 pips for 2 weeks, working on 6-7 pairs if i remember well. too good to be true.
i've been using it for 3 weeks now and my net profit is something between 15-20% up to date. but i still don't know how to use it.
as soon as i come with a winning strategy, i'll post it. it's just that it's gonna take more than i thought.
remember, this thread is a journal and i'll discuss my trades mostly, not my method/system.
Sorry to take more journal space , Scrat, but this is indeed a very important point. The bollinger (SMA) and linear regression deviations do NOT change
the previous values, whereas the parabolic regression channel may (I havent used it yet). Does anybody here (Scrat, Ozzie, others..) know of a particular advantage/logical reason for parabolic fitting to be better than linear or even a simple MA? I thought about this for a while and while the curve may be closer to each point of data (in this case the close values) the market does not have to obey polynomial equations of the form ax + bx^2 ..etc. Just my 2 cents on this, because if the parabolic line is giving you headaches from repainting AND it is not mathematically better, then maybe we could focus on the others.

PS: I have my E-U TP set at 1.3513, what about you?
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #119
  • May 21, 2007 12:47pm May 21, 2007 12:47pm
1. a regression is better than a MA because it gives you better entry points. when you draw a regression line/curve, you also get your standard deviations; the more data you use, the more reliable the curve is. still, you are shown data from the PAST, not future.
now, let's assume the price reaches 2 standard deviations from the curve; it doesn't mean it will go back to forever oscillate around a certain fixed regression line. it only means that past data suggests there are great chances for the price to go back to the regression curve.
2. the market does not obey any equation. let's keep in mind that regression curve is always changing and going towards or against the price (again, depending on past data). BUT they ALWAYS come together to a new cross again and again and again.
3. as for which one is better, i don't know. i'd like to say linear, but it's not always true. i've been using this formula for only a month, i won't know a lot until i bump into troubles .

Quoting Plutonite
Does anybody here (Scrat, Ozzie, others..) know of a particular advantage/logical reason for parabolic fitting to be better than linear or even a simple MA? I thought about this for a while and while the curve may be closer to each point of data (in this case the close values) the market does not have to obey polynomial equations of the form ax + bx^2 ..etc. Just my 2 cents on this, because if the parabolic line is giving you headaches from repainting AND it is not mathematically better, then maybe we could focus on the others.

PS: I have my E-U TP set at 1.3513, what about you?
your TP is ok for now.
Nihil Sine Deo.
  • #120
  • May 21, 2007 4:40pm May 21, 2007 4:40pm
i posted this pic into another thread, but i think it's relevant enough for my system.
1 year data used for regression, price at the lower limit. if it's going down, i doubt it'll be more than 1.0735 (give or take...).
i'm thinking if it's good to go long on this pair now...
Attached Image (click to enlarge)
Click to Enlarge Name: u-cad_1yr.jpg Size: 54 KB
Nihil Sine Deo.
0 traders viewing now