Pair Divergence Method | Page 48 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

What To Pair Divergence With? 14 replies

Divergence and hidden Divergence 122 replies

bo7a method... method for GBP/JPY 205 replies

Divergence method 11 replies

journal for pair divergence method 33 replies

Attachments: Pair Divergence Method
Exit Attachments

Pair Divergence Method

  • #941
  • Nov 25, 2010 12:31pm Nov 25, 2010 12:31pm
Quoting luxinterior
I hacked an OverlayChart indicator I had that displayed any sub pair so it displays as a line graph instead of a candle graph. I've also added an 'offset' variable that makes it easier to view the two symbols on the chart.

Code looks ugly but I threw it together quickly. It works fine.

Hope it helps some people.

Lux


Hello everyone, I am pretty new to FX trading and trying to learn and am making my way reading through this thread slowly trying to absorb it all. I would like to know as I am not overly skilled at MT4 (I can place trades and set SL and TP) how I can load and use this indicator of Luxinteriors so that I can use it to learn about and spot divergence.
I hope that the thread is not too cold and people are still using this method as it seems to have good sound principles.
I am still trading in demo and learning.
Thanks to anyone who may reply.
  • #942
  • Dec 13, 2010 3:06am Dec 13, 2010 3:06am
October was awful! 10 wins, 22 losses. Average loss was 81% of the average win. E = -0.21. Personal account lost 17.3%

November was great! 19 wins, 15 losses. Average loss was 0.74% of the average win. E = 0.22. Personal account was up 16.7%; however, I missed 7 trades (for various reasons) that I should have taken. I calculated a theoretical 24% return had you taken every trade, which would have made up for all of October's losses +3% or so.
Hindsight is 20/20
  • #943
  • Dec 15, 2010 4:32pm Dec 15, 2010 4:32pm
Every time I try to add the overlay indicator to the chart it tells me there is no data or the pair is invalid. Please help if you know what I'm doing wrong! Thanks.

Cory
  • #944
  • Dec 16, 2010 4:42am Dec 16, 2010 4:42am
Quoting Cbh123
Every time I try to add the overlay indicator to the chart it tells me there is no data or the pair is invalid. Please help if you know what I'm doing wrong! Thanks.

Cory
Open a chart corresponding to the pair you're trying to overlay. E.g. if you have an EURUSD 15m chart open, and you're trying to overlay USDCHF, open a new USDCHF 15m chart.
Hindsight is 20/20
  • #945
  • Dec 17, 2010 9:40pm Dec 17, 2010 9:40pm
Quoting cosgrove
Open a chart corresponding to the pair you're trying to overlay. E.g. if you have an EURUSD 15m chart open, and you're trying to overlay USDCHF, open a new USDCHF 15m chart.
Thanks cosgrove!

Can't believe I didn't think about that. I appreciate the help.

Cory
  • #946
  • Feb 2, 2011 10:22pm Feb 2, 2011 10:22pm
December stats (I only traded until 12/17 then took vacation until the new year. I did not go back and check how the system performed otherwise ):
9 wins 11 losses. Average loss was 75% of average win. E = 0.038

January stats (includes 1 win on MLK Jr holiday):
16 wins 15 losses. Average loss was 75% of average win. E = 0.097
Hindsight is 20/20
  • #947
  • Feb 23, 2011 9:56am Feb 23, 2011 9:56am
Thanks Cosgrove, - for keeping this thread up-to-date on your results.

This system does look very interesting. I like your SL to TP ratio, it helps to make up for losers.

I'm wondering (rhetorically) if most systems, in the long run, only return 50% winners & 50 % losers.

Prospector48
  • #948
  • Edited 11:08pm Feb 23, 2011 10:44pm | Edited 11:08pm
Quoting Prospector48
I'm wondering (rhetorically) if most systems, in the long run, only return 50% winners & 50 % losers.

Prospector48
I have had this thought recently as well, except I expanded it to postulate that as your R:R skews towards 0 or infinity, your % win rate will go towards 0% or 100%. For example, 10% win rate with a Risk:Reward of 1:10, or 90% win rate with a risk:reward ratio of 10:1 (numbers pulled out of a hat, obviously)

I'm not sure where the Random Walk/Efficient Market Hypothesis or real-life trading constraints (margin rules, etc) fit in there.

There are also many people who will argue that for your system statistics to mean anything, you need thousands upon thousands of trades, which I can somewhat agree to. If you go by that idea, then there is a non-trivial chance that the fundamental nature of whatever market you are trading may change before you reach that many system trades, thereby possibly invalidating all testing done so far. The next step in my mind would be to think "OK, so we're basically picking a system that seems to make sense and hoping for the best?" -- does that mean we're just hoping to "get lucky"?

Looking back just now, I have logged approximately 1000 trades with this system between April 2008 and today (only taken ones between April 2010 and today as live trades). Approximate win rate is 52.5%, and I use approximate here because I do not have TP=2xATR, SL=1.5xATR data for all of 2008-2009.

If you put those stats into the equity curve generator here: http://www.hquotes.com/tradehard/simulator.html (Win/Loss = 1.33, prob win = 0.525) the system looks very promising.
Hindsight is 20/20
  • #949
  • Mar 1, 2011 11:06pm Mar 1, 2011 11:06pm
February results! 15 wins, 14 losses, with average loss being 77% of the average win. Expectancy of 0.146.
Hindsight is 20/20
  • #950
  • Mar 8, 2011 3:35pm Mar 8, 2011 3:35pm
Quoting cosgrove
January stats (includes 1 win on MLK Jr holiday):
16 wins 15 losses. Average loss was 75% of average win. E = 0.097
Sorry, this expectancy should be 0.153 (E = (16/31)*1-(15/31)*0.75). It would be 0.125 without the MLK Jr day win. Didn't mean to mislead. March is going well so far!
Hindsight is 20/20
  • #951
  • Apr 5, 2011 4:49am Apr 5, 2011 4:49am
Theoretical March results:

14 wins
13 losses
Average loser was 75% of the average winner
Expectancy = 0.157

I missed a few trades this month and messed up another causing a loss instead of a win, so my personal results were not up to par with the system.
Hindsight is 20/20
  • #952
  • May 12, 2011 1:34am May 12, 2011 1:34am
April was quiet! I missed two wins, with one being on a bank holiday. I also had 7 losses in a row, which was pretty tough.

Theoretical results: 8 wins, 8 losses, with average loss being 75% of the average win. E = 0.125
Hindsight is 20/20
  • #953
  • Jun 7, 2011 2:49am Jun 7, 2011 2:49am
May was pretty active, with 14 wins, 22 losses. One loss occurred on a bank holiday, I missed a loss due to personal vacation, and I slept through a few entries as well. My personal account fared slightly better than theoretical results because of all that. Average loser was 73% of the average winner.

Expectancy = (% winners * average win) - (% losers * average loser) = 0.39*1-0.61*0.73 = -0.055
Hindsight is 20/20
  • #954
  • Aug 5, 2011 2:21pm Aug 5, 2011 2:21pm
I have temporarily ceased trading this method until EUR and CHF get back in line, if ever. Looking at taking it to AUD/USD and USD/CAD.
Hindsight is 20/20
  • #955
  • Aug 5, 2011 2:24pm Aug 5, 2011 2:24pm
Quoting cosgrove
I have had this thought recently as well, except I expanded it to hypothesize that as your R:R skews towards 0 or infinity, your % win rate will go towards 0% or 100%. For example, 10% win rate with a Risk:Reward of 1:10, or 90% win rate with a risk:reward ratio of 10:1 (numbers pulled out of a hat, obviously)

I'm not sure where the Random Walk/Efficient Market Hypothesis or real-life trading constraints (margin rules, etc) fit in there.
Power Law might have something to do with the quote. http://www.mercenarytrader.com/2011/...rket-behavior/
Hindsight is 20/20
  • #956
  • Oct 2, 2011 11:41pm Oct 2, 2011 11:41pm
Hi Cosgrove,

Do you trade the positive correlated pairs or only the negative.
I see some good setups on the 15 min charts.

ECO
Enter Signature
  • #957
  • Oct 3, 2011 11:06pm Oct 3, 2011 11:06pm
I was only trading EUR/USD vs. USD/CHF because they were the most correlated. I have stopped trading the method, but I have been meaning to look back at it once more since the SNB decision to hold EUR/CHF to a minimum of 1.2.

Correlation simply isn't enough though if you want a math proof for pairs trading.
Hindsight is 20/20
  • #958
  • Mar 1, 2014 11:49pm Mar 1, 2014 11:49pm
Quoting giraia_br
market is sideways.This is a trade you would take (if this bar close down)? EU 15 min {image}
Hello
CORRELATING PAIR DIVERGENCE -SIGNAL ONLY
I look at this picture right signal time!
I 1bar signal history!
What am I doing wrong?
Whether you assist me.
  • #959
  • Mar 2, 2014 3:39am Mar 2, 2014 3:39am
Attached File(s)
File Type: mq4 SecondChart.mq4   10 KB | 627 downloads
Attached File(s)
File Type: mq4 SecondChart.mq4   10 KB | 627 downloads
Quoting BenQ
{quote} ok, here's second chart courtesy of Serega Lykov. Place it on gbpjpy H4, for example, and enter eurjpy and 240 in the inputs and you'll have both on the same chart. I have a few others, and also one that does an inversion like you asked, but I have to dig those out. {file}
Can anyone modify this indi so that it accept the other pairs like DOW or S&P500 etc
MANY THANKS
  • #960
  • Edited 1:25pm Mar 2, 2014 9:40am | Edited 1:25pm
Quoting billbss
These backtests confirm what I concluded several months ago. This is one of the rare mechanical systems that is profitable. It may be the only one we'll ever see. It is also one I would not trade. I would not trade a system that shows a negative month. This method is a framework for a very profitable system. Anytime the basic method exceeds the expectation of random entries, it can be improved upon. I'm open to suggestions, particularly from Cosgrove and Artov and Lucid, who have worked with it.
billbss hi,

First of all, I must say that I really like your pragmatic approach in your inputs here on FF. Appreciate it!

There are several leading "indicators" which based on my experience in terms of Forex really work. I only use them as triggers mainly, as signals that is time to enter into a transaction.
One of these leading "indicators" is the Divergence on CII indicator (CII is one of the best Divergence indicators from my experience, possibly because is not normalized as other indicators ), other for example is the Slingshot from CCI (50,14 - you know better since I discovered it on your threads, thanks) , or the ADX contraction - expansion from forexalien (really good one, for entering and also for staying in the trade).

Having the Pairs divergence is a very good additional tool for announcing possible entries, in my view. Pairs correlation was always a subject which I believed have something of value, but things seemed complicated until now and you made them really simple and clear in your thread.

I have just ran into this thread of yours and I strongly believe you have something good here, thanks for this thread as well.
Will read it thoroughly this week, with great interest.

All the best to you,
Dragos
0 traders viewing now