Quick and Dirty EOD | Page 22 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

EOD trading system 1,037 replies

EOD -1 day hold 213 replies

Quicker and Dirtier EOD 117 replies

Best EOD strategy 5 replies

The meaning of EOD? 3 replies

Attachments: Quick and Dirty EOD
Exit Attachments

Quick and Dirty EOD

  • #421
  • Aug 2, 2012 11:32pm Aug 2, 2012 11:32pm
Quoting Wendavol
I am currently doing a manual backtest of the original system. This time the only different thing I am doing is replacing the monthly trend ema's. In its place, I am using the weekly candles. If the close of the last completed weekly candle is higher than the open of the candle before it, then I will look for long signals for the current week. Vice versa for shorts. If the close does not break the open of the previous candle, then there is no change in trend. By the way, in this test I am using Hugh's original entry and not the one I was testing...
Sounds interesting, thanks for posting.
  • #422
  • Aug 3, 2012 1:16am Aug 3, 2012 1:16am
Quoting Wendavol
I am currently doing a manual backtest of the original system. This time the only different thing I am doing is replacing the monthly trend ema's. In its place, I am using the weekly candles. If the close of the last completed weekly candle is higher than the open of the candle before it, then I will look for long signals for the current week. Vice versa for shorts. If the close does not break the open of the previous candle, then there is no change in trend. By the way, in this test I am using Hugh's original entry and not the one I was testing...
Are u using the 100SL 200TP rule or u use the min/max of the candle for SL?
  • #423
  • Aug 3, 2012 6:42am Aug 3, 2012 6:42am
Quoting lucariga
Are u using the 100SL 200TP rule or u use the min/max of the candle for SL?
I am using the min/max of the signal candle for SL with max SL of 100. My TP is fixed at 200.
  • #424
  • Aug 3, 2012 9:07am Aug 3, 2012 9:07am
Quoting Wendavol
I am using the min/max of the signal candle for SL with max SL of 100. My TP is fixed at 200.
Great... I'm impatient to see ur results.

Closed GBPAUD for 200 pips profit. This compensates the 200 pips of yesterday's loss...
  • #425
  • Aug 3, 2012 7:04pm Aug 3, 2012 7:04pm
what year is this? what i got from 2011 Jan till now is pathetic loser, which fits the logic of this system.

Quoting mahanxiong
Good job everyone.
My girl code this EA which's her first or 2ed EA.It's great.Welcome for any updates and please have a good deal with the chinese words in the code.Good luck.

Without MM:
http://www.forexfactory.com/attachme...1&d=1341199666

With MM:
http://www.forexfactory.com/attachme...1&d=1341199666

mahanxiong
  • #426
  • Aug 3, 2012 9:45pm Aug 3, 2012 9:45pm
Quoting charvo
what year is this? what i got from 2011 Jan till now is pathetic loser, which fits the logic of this system.
I do agree that something is wrong with the EA of that poster; I tested it and the results did not jibe with the original rules. What is it about this system that makes you think it is illogical? It is a simple trend following breakout system. It is also a work in progress. How would you improve or change this system to make it not such a pathetic loser?
  • #427
  • Aug 3, 2012 11:28pm Aug 3, 2012 11:28pm
Hi, please see the attached:

it is from Jan, 2011 to NOW. it may not be best quality test, but presenting a general idea that this is NOT that promising.

the original author's backtest might be a 10 yr history test. I'm not sure. although it looks like a holy grail, i'm sure that few can trade this since 2011 till now and still remain confident in this EA according to my test.

the problem of the current methodology is it contradicts itself. i personally don't believe that a trend-following EA may be successful with TP:SL = 2?????? the fundamental logic is "TREND FOLLOWING", and the method TP @ 200 pips? where's the following?

Only by this fact, I could guess 90% this current status is NOT working. Today i made some rough backtest suggesting additional 5% NOT working. that is, i guess it is 95% more likely to be a loser than a winner.

for a losing (imho 95% likely) system, it could be losing just a bit more than winning (which this system seems so). we still can improve it. the immediate way is to find a better exit----to stay longer with the trend. starting from some simple ways:
1. if use 5,8 MA for trend, then use something like 4X7 MAs to exit;
2. or enter 3 pos, exit 1st at 200 pips, exit 2nd at 300 pips, exit 3rd at 800 pips (who knows)
.......

all stupid idea did i present above, but at least that's the direction, imho.

i'm not being offensive, just being frank.

Quoting Wendavol
I tested it and the results did not jibe with the original rules. What is it about this system that makes you think it is illogical? It is a simple trend following breakout system. It is also a work in progress. How would you improve or change this system to make it not such a pathetic loser?
Attached Image (click to enlarge)
Click to Enlarge Name: TesterGraph.gif Size: 13 KB
  • #428
  • Aug 3, 2012 11:43pm Aug 3, 2012 11:43pm
this seems better, the logic is very simple to understand:

if a trend continues, the weekly bar must agree. and weekly bar suits better for entries of 100/200 pips than monthly MAs (to me).

if a trend reverse, the weekly bar reverse faster, but NOT for monthly MAs.

are you saying that you do the manual test using your weekly bars? or Hugh's original rules?

if getting promising results, plz post. it is very easy to code an EA for this method. i'm a little surprised that after so many pages, few paid good attention to some historical tests.......

Quoting Wendavol
I am currently doing a manual backtest of the original system. This time the only different thing I am doing is replacing the monthly trend ema's. In its place, I am using the weekly candles. If the close of the last completed weekly candle is higher than the open of the candle before it, then I will look for long signals for the current week. Vice versa for shorts. If the close does not break the open of the previous candle, then there is no change in trend. By the way, in this test I am using Hugh's original entry and not the one I was testing...
  • #429
  • Aug 3, 2012 11:57pm Aug 3, 2012 11:57pm
Quoting charvo
Hi, please see the attached:

it is from Jan, 2011 to NOW. it may not be best quality test, but presenting a general idea that this is NOT that promising.

the original author's backtest might be a 10 yr history test. I'm not sure. although it looks like a holy grail, i'm sure that few can trade this since 2011 till now and still remain confident in this EA according to my test.

the problem of the current methodology is it contradicts itself. i personally don't believe that a trend-following EA may be successful with TP:SL = 2??????...
I don't believe anyone participating in this thread is using the EA. I tested the EA and it does not agree with my manual backtesting. I don't consider those EA results as a basis for thinking this is a promising system. I am trying to contribute to this thread by testing different ideas. I am currently working on a different method for determining trend, as i think using the monthly MA's may be too large as we are looking to capture about 1.5 times the daily atr. So I have been doing a manual backtest which is taking a while because life gets in the way. So far results are good. 44% win rate and +1,624 pips on the e/u from July 2010 through June 2011.
  • #430
  • Aug 4, 2012 12:04am Aug 4, 2012 12:04am
Quoting charvo
this seems better, the logic is very simple to understand:

if a trend continues, the weekly bar must agree. and weekly bar suits better for entries of 100/200 pips than monthly MAs (to me).

if a trend reverse, the weekly bar reverse faster, but NOT for monthly MAs.

are you saying that you do the manual test using your weekly bars? or Hugh's original rules?

if getting promising results, plz post. it is very easy to code an EA for this method. i'm a little surprised that after so many pages, few paid good attention to some historical tests.......
I was also surprised that nobody posted any backtests. I did one with Hugh's original rules with result of +678 pips from August 2010 through June 2012. I then looked at the charts some more and started thinking of different entries, trend rules, etc. I started testing using an aggressive entry with an average of 5 times reward to risk, but some people here felt it was too risky so I stopped testing. So now I'm focusing on using the weekly candles as trend.
  • #431
  • Aug 4, 2012 2:35am Aug 4, 2012 2:35am
Another looser yesterday on CAD/CHF....
-100 for the week
  • #432
  • Aug 4, 2012 3:53am Aug 4, 2012 3:53am
Quoting Wendavol
I don't believe anyone participating in this thread is using the EA. I tested the EA and it does not agree with my manual backtesting. I don't consider those EA results as a basis for thinking this is a promising system. I am trying to contribute to this thread by testing different ideas. I am currently working on a different method for determining trend, as i think using the monthly MA's may be too large as we are looking to capture about 1.5 times the daily atr. So I have been doing a manual backtest which is taking a while because life gets in...
I looked at mahanxiong's code and I think it has some issues with the original rules. I recoded the EA to trade closer to the rules and ran a few tests over different time periods. I also didn't get very good results. Some discretion is needed, or we need to program more rules for the EA.

I could share the EA, but I used original code as a basis and I am not sure if the author agrees to share my changes of his code :-) . Will need to contact with him.

Julius
  • #433
  • Aug 4, 2012 7:50am Aug 4, 2012 7:50am
Quoting charvo
this seems better, the logic is very simple to understand:

if a trend continues, the weekly bar must agree. and weekly bar suits better for entries of 100/200 pips than monthly MAs (to me).

if a trend reverse, the weekly bar reverse faster, but NOT for monthly MAs.

are you saying that you do the manual test using your weekly bars? or Hugh's original rules?

if getting promising results, plz post. it is very easy to code an EA for this method. i'm a little surprised that after so many pages, few paid good attention to some historical tests.......
The EA at the start of the thread is flawed. It does not follow the original system rules. Also changes are already being made to target more pips from each trade while keeping the risk as small as possible. When I originally backtested the idea I performed extensive testing on the manual strategy tester and came up with acceptable results.
  • #434
  • Aug 4, 2012 8:52am Aug 4, 2012 8:52am
right, i've just spot that the EA trades differently from your rules. it is still a myth how the impressive results are obtained though.

your system has a great starting point, may be profitable after some "let it run" consideration are added. we will work on that.

Quoting Hugh Briss
The EA at the start of the thread is flawed. It does not follow the original system rules. Also changes are already being made to target more pips from each trade while keeping the risk as small as possible. When I originally backtested the idea I performed extensive testing on the manual strategy tester and came up with acceptable results.
  • #435
  • Aug 4, 2012 8:54am Aug 4, 2012 8:54am
great to know.

you know how much drawdown you experienced during the process? if the DD is too big compared to 678, then it will still pose doubt to the profitability.

Quoting Wendavol
I did one with Hugh's original rules with result of +678 pips from August 2010 through June 2012.
  • #436
  • Aug 4, 2012 11:43am Aug 4, 2012 11:43am
Quoting charvo
great to know.

you know how much drawdown you experienced during the process? if the DD is too big compared to 678, then it will still pose doubt to the profitability.
It's been a while since I did the backtest so I'm not sure exactly what it was but it would have been acceptable or I wouldn't have continued with my theory.

You ask a lot of questions about backtesting for someone who backtesting means nothing to
  • #437
  • Aug 4, 2012 1:15pm Aug 4, 2012 1:15pm
T
Quoting charvo
Hi, please see the attached:

it is from Jan, 2011 to NOW. it may not be best quality test, but presenting a general idea that this is NOT that promising.

the original author's backtest might be a 10 yr history test. I'm not sure. although it looks like a holy grail, i'm sure that few can trade this since 2011 till now and still remain confident in this EA according to my test.

the problem of the current methodology is it contradicts itself. i personally don't believe that a trend-following EA may be successful with TP:SL = 2??????...

Charvo,

Not sure what your goal is here.

"the original author's backtest might be a 10 yr history test. I'm not sure. although it looks like a holy grail, i'm sure that few can trade this since 2011 till now and still remain confident in this EA according to my test."

The original author and the EA have nothing to do with one another. To be honest, I'd totally forgotten about the posts about the EA.

the problem of the current methodology is it contradicts itself...the fundamental logic is "TREND FOLLOWING", and the method TP @ 200 pips? where's the following?

The method of entry is "trend following". We're entering on the daily in line with (following) the trend of the monthly.

Only by this fact, I could guess 90% this current status is NOT working.

You Guess!!! I personally don't want to guess at anything. Do you have any data to back up your guess?

we still can improve it. the immediate way is to find a better exit

Now here I agree with you, as has already and continues to be discussed on the thread.

I certainly appreciate anyone's constructive thoughts on ways to improve this or any other system. But it's hard to see the benefit when someone starts out by bashing the thread and OP.
  • #438
  • Aug 4, 2012 9:30pm Aug 4, 2012 9:30pm
Charvo,

Not sure what your goal is here.

you don't need to be sure since it is not for you

"the original author's backtest might be a 10 yr history test. I'm not sure. although it looks like a holy grail, i'm sure that few can trade this since 2011 till now and still remain confident in this EA according to my test."

The original author and the EA have nothing to do with one another. To be honest, I'd totally forgotten about the posts about the EA.

if u'd like to misunderstand "the original author" as of this thread instead of plain simply of the EA, feel free.

the problem of the current methodology is it contradicts itself...the fundamental logic is "TREND FOLLOWING", and the method TP @ 200 pips? where's the following?

The method of entry is "trend following". We're entering on the daily in line with (following) the trend of the monthly.

the entry is following, but the position is NOT really following(if tp @ 200pip), isn't it?

Only by this fact, I could guess 90% this current status is NOT working.

You Guess!!! I personally don't want to guess at anything. Do you have any data to back up your guess?

i showed my data, and being modest to say it is 90%. or you show me ur data that it is profitable??? on the other hand, everytime i place a trade, its direction represents my guess. anyone who trades doesn't guess???

we still can improve it. the immediate way is to find a better exit

Now here I agree with you, as has already and continues to be discussed on the thread.

I certainly appreciate anyone's constructive thoughts on ways to improve this or any other system. But it's hard to see the benefit when someone starts out by bashing the thread and OP.

i'm not bashing, i've said, being frank not offensive. trading's nature is very objective and cruel. being over polite or considerate is never the path to the goal. if something is wrong, say it is (probably) wrong; or right vice versa. i believe that i've discussed problem of the methodology instead of any person
  • #439
  • Aug 5, 2012 10:05am Aug 5, 2012 10:05am
Charvo - the reason you are angering people in this thread, is because you called the strategy and the logic of the OP 'pathetic', but then didnt give much info about why you think the strategy wont work. As a trader with 12 years of experience, as well as job experience working at the CBOE in Chicago, there is nothing pathetic about either this strategy or the logic behind it. We are looking for an edge, using a breakout that has a bias in the favor of the trend, with possibly using a trailing take profit to let the profits run. I agree that the 200-pip TP is arbitrary, but that is what we are experimenting with (a better TP would be to base it on a Round Number or a S/R point...but I believe in testing ideas one at a time, instead of bouncing around from idea to idea without getting serious data about how it performs, which is one of the problems with forums like this...data overload, and too many "cooks in the kitchen" and then many good ideas never get a chance to be tested completely before they are discarded).

Also, what "proof" do you have to make you think that the method wont work? I See that you posted a screen shot of the output of an EA. Which version did you test? The 100/200 version, or the version where the SL is just beyond the end of he candle? And I saw that you posted only one year of data. Was that just for one currency pair? Maybe other pairs were wildly profitable during that same time period. And then to top it all off, you claim in your profile that backtesting is worthless to you. So, why are you bothering with that? And by the way, there are multiple people "forward-testing" this, with live capital.
  • #440
  • Edited 11:00am Aug 5, 2012 10:38am | Edited 11:00am
One more issue with "BackTesting" using MT..... unless you have high quality data, and are an expert programer, the results of Mt back testing is suspect... not to mention the nature of MT being suspect anyway............ I would prefer to see backtest results from a platform with a little more integrity such as trade station or trading blox for example.

I however am fluent in neither and do my back testing the old fashioned way.... a short period with pencil and paper then live forward test from there.

At this point I do not have much data since I just got started really looking at this system. However with that said I am encouraged and as I have mentioned earlier in this thread I use a very similar signal to add to my trend following positions and that works fine.

JMHO

As to traders using a live account and forward testing there is one here who has been trading this for just over a month, he is using what I consider a conservative leverage, he is following the rules, he is trading multiple pairs and he is up over 24% ..... I think that is a pretty good testimony to the system........


.
1 trader viewing now