GBP/USD - one trade a day | Page 19 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

EUR/GBP, EUR/USD, GBP/USD, AUD/USD, CAD, NZD, JPY, and more... 15 replies

GBP/AUD, GBP/CAD, GBP/CHF and GBP/NZD thread 126 replies

ACD Method on GBP/JPY One trade a day 451 replies

The best way to trade is one trade a day 76 replies

two version of FF tunnel Chart, one say down, one say up, which one is correct? 3 replies

Attachments: GBP/USD - one trade a day
Exit Attachments

GBP/USD - one trade a day

  • #361
  • Feb 15, 2011 7:02am Feb 15, 2011 7:02am
Quoting Chris_B
Said in a few words:

1. I watch 01:00 candle to pick direction instead of 00:00 candle

2. I don't recalculate my target every day...so I don't close the current trade if the target is reached but I leave it open until the end of the candle, this may seem a slight difference but it's not...because for example...this month i'm 77 pips in gain...the original strategy would put my target at only 3 pips tomorrow...my strategy doesn't put that limit, I let the trade close by itself...if it makes 50 pips for example..I've made a good amount of pips...
Can you please check your Feb 4th trade, because it was stopped out as entry was at 6147.3 and price went up to 6170.4 just after the entry.

Thanks
  • #362
  • Feb 15, 2011 7:26am Feb 15, 2011 7:26am
Quoting Blackeagle
Can you please check your Feb 4th trade, because it was stopped out as entry was at 6147.3 and price went up to 6170.4 just after the entry.

Thanks
I don't know why...but I have different numbers on my feed....
I will check if there are others differencies...
Attached Image (click to enlarge)
Click to Enlarge Name: feb4.JPG Size: 63 KB
Market is like a puzzle, YOU have to fit the pieces!!
  • #363
  • Feb 15, 2011 8:04am Feb 15, 2011 8:04am
Hi cg + all,

I have spent a reasonable amount of time backtesting and tweaking this system over the last few weeks I think ~60 hours. (I am an IT worker currently between contracts) I have written an EA to help me with this, and have been testing on data back to Jan 2007.

I have changed the system as proposed by cgrey. I wont post what I have done here out of respect, however I will post the few things I have learned and you can draw from this your own conclusions. Please understand that I have modified the system to fit with my view of risk / reward and the below may or may not apply to the system as proposed by cgrey. I hesitated to post this but think it may move the system forward for all. These findings are based on average returns from jan 2007.

1) I have seen the need to aim for more than 10% per month as big gains are needed to knock out the losing months.

2) Some months will run and run into the red - I have decided to limit my drawdown, and not take any more trades after a given % loss.

3) There are huge possible gains in given months eg I have found +60% is possible at least once per year. The question is, how do we identify and keep pursuing such months but still bank our gains in average months?

4) The day of the week has an effect on long term gains. For some reason, and I can't explain this, the system doesn't seem to work long term on wednesdays! It doesn't make any sense to me but here are my findings with modified MM to allow the tests to run long term:

friday 1.3 thurs 28.1 Wed -50 tues 41.3 mon 60.2
I think it would be worth someone with a different perspective to me (I am running these tests completely automated) investigating if I am right or wrong with this one.

5) I don't have much faith currently in the use of the gap filter I haven't found it to be useful long term, my experience is that when we don't trade everyday we reduce our chance to meet the target consistently. One good day can turn around a whole losing month and by filtering trades we can miss these 'big days'.

6) Most losing days the SL is hit within the first few minutes, I don't know how to overcome this. Looking at the charts we would have a good chance to pick these out but due to the nature of this system and the fact that I am planning to trade by EA this isn't a good solution. I have an idea to split the lots across the open eg at 7:55, 8:05, 8:15 etc. I haven't done any testing but will post again if I find anything interesting.

7) Winning trades can often be left to run longer than 1 hour. Again draw your own conclusion from this.

------------------------------

I am not trying to derail this thread or change the underlying system as proposed, I am just posting to flag up a few details that have found had an effect on MY backtesting. The results and suggestions I have may well not hold true for the future but I post them honestly to show my current expectation for the system.

I am very wary about projecting results based on backtests (Especially as I have seen small changes to the system can give similar results only for one year to fall of the edge of a cliff.) But on the results from jan 2007 I have 2 profiles emerging. One is a lower risk allowing a 5% DD per month and showing 2.1% account growth per month averaged over the last four years. The other which departs more radically from the system shows an averaged 9.1% gain per month but often suffers huge drawdowns of ~60%. I am very aware that results from backtesting show us only a small amount of the picture but it is my opinion this system give us an edge, an increase in probability or whatever else you want to call it. How big this advantage is would have to be proved I believe by forward testing.

I know there is an EA already produced for this system that many may be using, but the EA I have written is fairly flexible and as such I would be happy to change parts and run tests to improve this system based on suggestions from the community. That is of course providing this is acceptable to cgrey and the changes are within my knowledge of Mql4.


dave.
  • #364
  • Edited 8:27am Feb 15, 2011 8:08am | Edited 8:27am
Hi dirtytough,

Quoting dirtytough
I do think you should incorporate the rule of not taking the trade if the gap is 9 pips or less.
This is looking more and more likely to be incorporated in to my set of rules, at least. Others can choose whether or not they want to set this filter. So far in Feburary, it would have filtered only one trade, but as it happens, that was a trade that stopped out. Other months that I've looked at this manually have all seemed to benefit from it, either in getting to target sooner and/or having less drawdown.

Quoting dirtytough
Another question for the guys who trade this manually. What if you didn't enter your trade right away and waited to put the trade in a little closer to your sl? You would have a better r/r that way.
I noticed someone else had answered this, and exactly the same way I would have... If the trade takes off in our direction as soon as we enter and never looks back, we would never be in it if we're waiting for it to come back down.

Quoting dirtytough
One more idea for all the traders that are thinking of trading this for more then 10% a month. Set your tp at 10% like cgrey does. Once you make 10% on however many trades it takes to get there, reset and use 10% as your tp again. Same if you have a dd of 10% reset and start over on your 10% goal. I think it would keep your tp realistic that way.
I don't see anything wrong with this approach. Haven't personally tested it either. I use the monthly "boundary", again, going back to the perspective of a business. My business accounting is done monthly, so this just kind of evolved from that.

Quoting dirtytough
Also cgrey, as you can see this is my first post and I would like to ask you about a mm strategy for a trade that I use...
Welcome. Even if you had enough posts for PMs, mine are turned off, so you wouldn't have reached me that way.. One recommendation as far as MM goes, and of course, this is my personal belief and others may tell you differently - never take a trade with a higher risk than reward. It will bite you in the end. {here comes a soapbox moment....} So many "systems" here have such broken money management, it's no wonder 95% of all retail traders fail. And some of them are fundamentally and/or technically sound as far as entries and exits go. No matter how many winners they get as opposed to losers, the first string of a few losers in a row puts you in a hole that you will never get out of. It's usually after a few weeks - maybe even a couple of months... Things are moving along nicely in a trending market. "We have 80% winning trades, so risking 2-3 times what we get is not a problem". Then the string of losses start, and three losses just wiped out the last 10 wins. A few more losses come, and now it takes 10 wins just to get back to break even. Magically, the thread then dies out... There are probably at least 100 stories like this here, all of which were systems that probably could have been at least somewhat profitable with proper money management. And this comes from experience... I had a system a few years ago that I felt exactly the same way about - HORRIBLE risk:reward, but I had lots of winners, so I was sure it didn't matter. And when some more experienced peeople pointed this out, I was sure I knew more than they possibly could about my own system, and ignored them. After a couple of months, the string of losers that came made me realize just how right they were...
  • #365
  • Feb 15, 2011 8:12am Feb 15, 2011 8:12am
Quoting puntalara
Nice day and I nearly did not take it as price was very close to the opening price on the 00.00 candle

About 15 pips I think
Personally, I think the 9-pip threshold that roundrock suggested is the right number if you plan to filter based on the distance between the two opening prices. Glad you didn't miss ths one
  • #366
  • Feb 15, 2011 8:20am Feb 15, 2011 8:20am
Quoting Chris_B
Said in a few words:
Hi Chris,

I'm going to need to spend some time taking a closer look... Some of your numbers are WAY different than mine in Februray. Even on trades that we would have both entered the same way based on comparing 00:00 or 01:00 to 08:00, you have a couple of winners when I had losers, or the pips won were much different when we both had winners.

I only took a very quick look, but it seemed at first glance that there was only one trade so far this month that we would have entered in opposite directions (because of the difference in the 00:00 and 01:00), you would have taken a winner there, where I lost. But it doesn't make up the net difference in pips - you're currently +77, I'm currently +3

Since we haven't been above the 10% threshold yet, your leaving that "last" trade open beyond the target shouldn't have come in to play yet, either..

I need to dig in to your results further, but I'm wondering if you're doing anything at all differently with closing trades (either sooner or later than 1-hour if TP or SL wasn't hit)??
  • #367
  • Feb 15, 2011 8:52am Feb 15, 2011 8:52am
Quoting dave0
Hi cg + all,

I have spent a reasonable amount of time backtesting and tweaking this system over the last few weeks I think ~60 hours. (I am an IT worker currently between contracts) I have written an EA to help me with this, and have been testing on data back to Jan 2007....
Excellent stuff Dave... This is exactly what I wanted to get out of this - to present an idea that people can make work for them. I like that you put some serious time in to this to get it to a place you're comfortable with.

I encourage discussion of your thoughts on how to improve the general concept. Of your list, I've thought a lot about #7 myself, but haven't found a good way to deal with it (yet). Of course, I'm against giving anything back once I have banked it, so that limits my options Similar thought about why I haven't pursued #3 further either.

My only request, should you choose to make your EA public, is that you open a new thread to direct people to it. I would prefer this thread stay clear of all of the issues that usually pop up as soon as an EA is made available - "it doesn't work for me", "can you add this...", "how do I set the time zone", "why are my results different than yours" etc, etc...

Congrats on the improvements you've made. I'm sure there are people looking to do more with this than how I've presented it, and will be happy to hear how you've done that.
  • #368
  • Feb 15, 2011 9:07am Feb 15, 2011 9:07am
Quoting cgrey
Hi Chris,

I'm going to need to spend some time taking a closer look......
No...I don't do anything different...the trade opens at the beginning of the candle and closes at the end of the candle...
Anyway...I don't consider spread and this...believe me...can make a big difference....and then it seems that my data feed is little different...this may be another reason for the different result.

Today I don't have the time..but tomorrow I will make a comparison between my Strategy Trader feed and my Metatrader feed, in order to understand where are these differences...

I don't really want to present here results that are completely out of reality...
Market is like a puzzle, YOU have to fit the pieces!!
  • #369
  • Feb 15, 2011 9:10am Feb 15, 2011 9:10am
Quoting Chris_B
I don't really want to present here results that are completely out of reality...
It may be mine that are out of reality

When I have some time over the next couple of days, I'll try to pinpoint some of the specific differences I've seen - that may help.
  • #370
  • Feb 15, 2011 9:19am Feb 15, 2011 9:19am
Hello everyone, hope all is well. In a matter of 2 weeks we have seen the ugly side and the beautiful side to this strategy. It's important to note however that both sides could play out longer. We got down 9% early, but as has been said before, the hole could of gotten deeper. Conversely, the gains which have made up the 9% loss could have been greater or can continue. The point I'm trying to make is even though we are booking results monthly, this is a long term strategy. I think it's prudent to not get too wrapped up in one month's results. However, this one looks like a winner.

Chris's method is in interesting with regards to letting the last candle close and not at take profit. I think this is where it depends on what kind of trader you are. If you are like cgrey and looking to hit 10% and close for the month, then why risk it getting to your target and then reversing. There I think you should close out and book the month. If you are going to reset your 10% and continue the next day, then I think letting it run makes sense and may be worth the shot. I think I'm more like Chris in terms of continuing the trading every day even if target reached before month end, however, I have a hybrid method for that last candle. It kind of takes both cgrey's and Chris's style and combines them (sort of). Lets say the target is 3 pips to make 10%, instead of closing out when candle is done (9 GMT), I would double the target. In this case TP = 6. I would do this only when the target is less than 10 pips (1-9 pips). Just a thought.
  • #371
  • Feb 15, 2011 10:19am Feb 15, 2011 10:19am
some excellent posts today and very intersting observations.

abour entering on a retracement closer to SL, i did backtesting using EA, it gave worse results than regular entry. best example is today. price didnt even retrace a pip and got us 60 pips in the end

i am now 'scared' of the wednesday trade (tomorrow).

has anyone spend time on another pair and other entry candle hours ?

very rarely u see something works on GU bit not EU
  • #372
  • Feb 15, 2011 10:30am Feb 15, 2011 10:30am
Quoting Blackeagle
What if price goes in trade direction without any drawdown as soon as the position is opened (as was the case today)?

Hope this helps.
Yep I thought of that also. And thought if a backtest shows it dd 5 pips 95% of the time before it go's positive it might be worth implementing. Probably wouldn't help though like you said.


Not trying to change anybodies system, just throwing out an idea that might offer less risk.
  • #373
  • Feb 15, 2011 10:35am Feb 15, 2011 10:35am
Quoting cgrey
Hi dirtytough,


.
Thanks for taking the time to respond. Like I said I read through some of your old journals. I noticed you used to have a really bad r/r on your system 1 trades. I used to think it didn't matter also but am learning that it does. Looks like you figured out how to trade successfully so congrats.

Thanks for the advice cgrey and everybody else in this thread. I am learning from all of it.
  • #374
  • Feb 15, 2011 11:09am Feb 15, 2011 11:09am
Quoting dirtytough
Thanks for taking the time to respond. Like I said I read through some of your old journals. I noticed you used to have a really bad r/r on your system 1 trades. I used to think it didn't matter also but am learning that it does. Looks like you figured out how to trade successfully so congrats.

Thanks for the advice cgrey and everybody else in this thread. I am learning from all of it.
Those were the days...

It's nice when you can look back several years later and see just how far you've come.

Hang in there. Learning is key. Put your mind to it - develop a plan and stick to it, and you will be successful. Trading isn't rocket science (though, don't be fooled in to thinking it is simple, either). But I do believe that anyone that dedicates themselves to learning their business can succeed.
  • #375
  • Edited 12:02pm Feb 15, 2011 11:48am | Edited 12:02pm
Quoting dave0
Hi cg + all,

I have spent a reasonable amount of time backtesting and tweaking this system over the last few weeks I think ~60 hours. (I am an IT worker currently between contracts) I have written an EA to help me with this, and have been testing on data back to Jan 2007.
Good stuff here Dave. Since the market is always changing, it can't be solved like a puzzle, but the more familiar we are with it the more likely we are to succeed.

Quoting dave0

1) I have seen the need to aim for more than 10% per month as big gains are needed to knock out the losing months.
This is currently the biggest question I have with this system. Is it profitable (or most profitable) in the long run only shooting for 10% then stopping. What is the best way to get the most out of this system in a month. In my opinion, if the mm is good enough to shoot for 10% then why stop there? The big question for me is at what target do I want to recalculate the next target AND do we stop and recalculate when in drawdown at some point (which is something you point out)?

Quoting dave0

3) There are huge possible gains in given months eg I have found +60% is possible at least once per year. The question is, how do we identify and keep pursuing such months but still bank our gains in average months?
I would trade each day and recalculate target after every 10% target reached. That is my personal preference. What do you think?

Quoting dave0
4) The day of the week has an effect on long term gains. For some reason, and I can't explain this, the system doesn't seem to work long term on wednesdays! It doesn't make any sense to me but here are my findings with modified MM to allow the tests to run long term:...
Surprised to see this. Don't know what to think. I guess I would start by trading on Wednesday and if I notice what has been the case in the past then just stop trading on wednesday.

Quoting dave0

5) I don't have much faith currently in the use of the gap filter I haven't found it to be useful long term, my experience is that when we don't trade everyday we reduce our chance to meet the target consistently. One good day can turn around a whole losing month and by filtering trades we can miss these 'big days'.
I'm inclined to always enter a trade, even when the gap is small. The most I have to lose is 20 pips, but I have the possibility of getting more than 20+ pips. Unless the evidence becomes overwhelming, I will continue to trade on small gaps.

Quoting dave0

6) Most losing days the SL is hit within the first few minutes, I don't know how to overcome this. Looking at the charts we would have a good chance to pick these out but due to the nature of this system and the fact that I am planning to trade by EA this isn't a good solution. I have an idea to split the lots across the open eg at 7:55, 8:05, 8:15 etc. I haven't done any testing but will post again if I find anything interesting.
There is no way to overcome this. You could try staggering entries, but if price shoots in your favor right from the entry then you miss out on entering the rest of your order. Sometimes you just got to s**t in your shorts and jump in the pool.

Quoting dave0


7) Winning trades can often be left to run longer than 1 hour. Again draw your own conclusion from this.
In my opinion, the trade could go for you or against you after one hour so to me this is negligible. For simplicity's sake and ease of trading I would just close the trade after one hour. You should have already gotten the big wave anyway (at least on most days).
  • #376
  • Feb 16, 2011 2:57am Feb 16, 2011 2:57am
Hi guys!

I've seen some interesting post in the thread in the last hours...

For what I can say...I've made a serie of backtests and as I said I found the best results in what I'm doing now....this system ain't broken...so don't fix it too much.

The concept is a good idea and can bring to you an average of 10% every month...
Surely...some month you can go over 10% if you don't stop...but I've seen more than a few months in my backtest where you got your 10% and then finish the month losing....
Are you really not happy with a 10% monthly?? Even some professional traders can't do this percentage consistently....

GBPUSD is the cross that works...and obviously there is a reason...and it's because of London open...in London they trade GBP...this is the reason..maybe you can try GBPJPY but I don't think other currencies will work (EURUSD is less volaitle).

I've also optimized with others hours to take the trade and....these are the best results (for me):

take the trade at the beginning of 08:00 GMT candle and make comparison with 01:00 GMT candle.
Even original 00:00 GMT candle gives good results but the drawdown is much higher (by my backtest)
Market is like a puzzle, YOU have to fit the pieces!!
  • #377
  • Feb 16, 2011 3:08am Feb 16, 2011 3:08am
A (little) win...+6 pips
  • #378
  • Feb 16, 2011 3:10am Feb 16, 2011 3:10am
This is today trade...+8,4 pips (without spread)

Does it match with yours??

Open at 1.61762
Close at 1.61678

My platform says that in February i'm at +84 pips...so target is reached and slightly exceeded.

Today I will make comparison with MT4 feed and will add spread to my backtests...
Attached Image (click to enlarge)
Click to Enlarge Name: today.JPG Size: 65 KB
Market is like a puzzle, YOU have to fit the pieces!!
  • #379
  • Feb 16, 2011 3:15am Feb 16, 2011 3:15am
Quoting Chris_B
Hi guys!

I've seen some interesting post in the thread in the last hours...

For what I can say...I've made a serie of backtests and as I said I found the best results in what I'm doing now....this system ain't broken...so don't fix it too much.

The concept is a good idea and can bring to you an average of 10% every month...
Surely...some month you can go over 10% if you don't stop...but I've seen more than a few months in my backtest where you got your 10% and then finish the month losing....
Are you really not happy with a 10% monthly?? Even...
Hi Chris,

Thanks for your evaluation...but I am puzzled. In looking back several trades, there doesn't seem to be much if any difference in comparing the 08:00 GMT candle to either the 00:00 candle or 01:00 candle...there's only 1 hour difference. Maybe it serves to change trade direction entries for the better on some of the really close price comparisons? Am I missing something here?

Thanks,
BC2
  • #380
  • Feb 16, 2011 3:17am Feb 16, 2011 3:17am
Quoting Bobcat2
Hi Chris,

Thanks for your evaluation...but I am puzzled. In looking back several trades, there doesn't seem to be much if any difference in comparing the 08:00 GMT candle to either the 00:00 candle or 01:00 candle...there's only 1 hour difference. Maybe it serves to change trade direction entries for the better on some of the really close price comparisons? Am I missing something here?

Thanks,
BC2
It's just an optimization over the last 18 months...01:00 candle gives better results in the long run...maybe not much difference in just 1 month and yes..the only difference can be the direction of the trade...
Market is like a puzzle, YOU have to fit the pieces!!
1 trader viewing now