The grid hedge - the core version | Page 7 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Flexy Grid EA (Grid Trading) 1,363 replies

Grid - New indicator that replaces chart grid 60 replies

Grid (not grid system) or horizontal lines indicator 6 replies

Grid, Burst Grid, Grid Risk and Grid Profit 1 reply

Double Time Grid and Pip Grid 0 replies

Attachments: The grid hedge - the core version
Exit Attachments

The grid hedge - the core version

  • #121
  • Apr 3, 2008 6:12am Apr 3, 2008 6:12am
ISY, If I were to close my D and E positions I would be closing two longs. Once I do this I am 3-3 whch means my whole position is locked unless we get a big move to retrigger the new orders. Seems better off to take the bullet. I have decided to put SL's on my longs. If they get hit and my two shorts get triggered I make it back fast. If they get hit and my shorts stall I will sell it all. So, basically I only need a small downside run to get it back. If I have a down and up I take the 300 or so loss.
  • #122
  • Edited 6:53am Apr 3, 2008 6:51am | Edited 6:53am
Hi. I tested the EA this morning. At first all seemed ok as there were 10 orders in total set. Then all of a sudden 300+ more pending orders were created, of which 100+ had the same sellstop entry. Within a minute i found myself with 116 sells activated at the same price . This was on my live account, which i know is a bit stupid to test the EA on.. but i kept the lotsizes extremely small..

Fortunately, i managed to close all of them for a profit of 4 pips each.

Anyway, thought i give you guys head up on this one..
Would be great if the EA worked though..
Liberate me ex inferis
  • #123
  • Edited 8:50am Apr 3, 2008 8:11am | Edited 8:50am
Not sure what would cause the EA to continue opening trades like that. Again I am not a coder, therefore I would appreciate it if a coder could refine the EA. Definitely not a good idea to test this EA on live acct.

TimeFreedom

Quoting Ronyn76
Hi. I tested the EA this morning. At first all seemed ok as there were 10 orders in total set. Then all of a sudden 300+ more pending orders were created, of which 100+ had the same sellstop entry. Within a minute i found myself with 116 sells activated at the same price . This was on my live account, which i know is a bit stupid to test the EA on.. but i kept the lotsizes extremely small..

Fortunately, i managed to close all of them for a profit of 4 pips each.

Anyway, thought i give you guys head up on this one..
Would be great if the EA worked though..
  • #124
  • Apr 3, 2008 8:12am Apr 3, 2008 8:12am
hey rony..... there needs to be an additional switch so the ea can tell if orders have already been placed today.... otherwise on each tick, new orders will be placed cause there is nothing saying "stop once" the first round was placed.... those following will be placed with the most current bid/ask information......


the ea will placed trades soon after it's attached to the chart.... if a certain time of day to place the trades is desired, that should be added also..... along with a 'close at' time, if so desired.....h
to trade and code, keep both simple... no call to impress....h
  • #125
  • Apr 3, 2008 8:27am Apr 3, 2008 8:27am
Quoting Abnormalone
Here is a spreadsheet using FXDD history. It appears that FXDD's day starts 2 hours earlier than your brokers and the spread is 2 pips. For the test I made first entry + or - 10 pips from the Open price.
From 1/2/2007 - 4/1/2008 total loss -1704 pips.
This is very important information. Has anyone double checked this?
  • #126
  • Apr 3, 2008 8:45am Apr 3, 2008 8:45am
Here is an updated HedgeGrid_v1.01 EA, which should not open multiple orders after the initial 5 trades. I have tested this new version on a demo by switching between chart timeframe and restarting trading platform and it did not open unwanted additional trades.

I also included a startGrid option, which tells the EA to start the grid once the setting is set to true. This is to allow you to watch the charts with the EA attached to the charts and when you are ready to start a grid then set startGrid to true.

Note: Do not use in live acct. until we have all the bugs addressed.

TimeFreedom

Quoting hayseed
hey rony..... there needs to be an additional switch so the ea can tell if orders have already been placed today.... otherwise on each tick, new orders will be placed cause there is nothing saying "stop once" the first round was placed.... those following will be placed with the most current bid/ask information......


the ea will placed trades soon after it's attached to the chart.... if a certain time of day to place the trades is desired, that should be added also..... along with a 'close at' time, if so desired.....h
Attached File(s)
File Type: mq4 [ea]_GridHedge_v1.01.mq4   15 KB | 932 downloads
  • #127
  • Apr 3, 2008 10:13am Apr 3, 2008 10:13am
Quoting jeta
This is very important information. Has anyone double checked this?
The spreadsheet assumed you would close all entries and start over again each day.
If you let your good entries run for several days (which is what isydeb does) it would make a real big difference in the results.
  • #128
  • Apr 3, 2008 10:20am Apr 3, 2008 10:20am
TF,

Did you backtest the new EA? Im wondering if 5 orders is the best. Also, wondering about spacing options?

BTW, I closed all the trades this morning and ate the loss. Will start over wither tonight or next week. MAy not enter tonight due to the NFP
  • #129
  • Apr 3, 2008 11:06am Apr 3, 2008 11:06am
Quoting scottyb159
Qman,

I believe the closing out method was meant for when you were in a profit overall, not a loss. I could be wrong but that was how I read it.
Hey scottyb159, here is the quote of Advance Technique #1:

"If you are able to babysit your trades or monitor them on an hourly basis well then this is a technique for you. When in profit, be it on the short side or the long side, close the losing positions and set a stoploss to the winning positions at the breakeven of the last winning position. And place other set of orders on both sides. Max 5 trades per side again. By doing this you are locking in profit and adding to a winning position but still controlling your risk. You can do this on an hourly basis. And if you're lucky you can ride a trend of days without ever increasing your risk."

Right now I can't be sure to check status every hour... especially the London session (sleep may be over-rated, but I have enough illusions as a rested trader -- I don't need to add halucinations to my problems). So, I'm thinking of using this advanced technique, but with a twist. Something like close losing positions at the start of every session, lock the profit on the winners, and reestablish the grid. I might be able to do this and one or two other adjustments during a day. I think this would have made a killing this week. I also think it would have saved your bacon over the last day or so.

I am still just a beginning programmer, but I'm trying to create an EA that will do this, allowing you to set times for reanalysis, down to the minute. I'm trying to make it aware of the trades the EA has placed in the past, so that you can turn it off, change analysis times (probably with an eye to scheduled news) and restart it. You could then make adjustments, even to the grid size and spacing at your own discression, even multiple times throughout the day. It will otherwise follow the rules in this advanced technique. It will close the losing positions, and it will change the SL on the winning ones, then create a new grid.

I can't get stuff like this done overnight. I'm just not good enough. I do have a start, and will steal some code from the EA TimeFreedom posted to jumpstart my efforts. Warning you in advance though, if I get this done and you all want improvements to it, the turnaround will be slow. Having said all that, if someone else is working this up, do let me know. I've got other things I can throw my limited time at.
  • #130
  • Apr 3, 2008 12:24pm Apr 3, 2008 12:24pm
There are many ready made grid Experts for sale on the internet...are they worth it?
  • #131
  • Apr 3, 2008 12:33pm Apr 3, 2008 12:33pm
Qman,

As you quoted, that technique (I belive ) is when there is first an OVERALL profit. In my mind it would not work if not already in an overall proft, because that is the profit you are really locking in when tyou get rid of the losing trades. ISY, correct me if I am wrong?
  • #132
  • Apr 3, 2008 12:38pm Apr 3, 2008 12:38pm
Quoting Abnormalone
The spreadsheet assumed you would close all entries and start over again each day.
If you let your good entries run for several days (which is what isydeb does) it would make a real big difference in the results.

Have you done the backtest for that? Stats?
  • #133
  • Apr 3, 2008 1:32pm Apr 3, 2008 1:32pm
Quoting scottyb159
Qman,

As you quoted, that technique (I belive ) is when there is first an OVERALL profit. In my mind it would not work if not already in an overall proft, because that is the profit you are really locking in when tyou get rid of the losing trades. ISY, correct me if I am wrong?
Scottyb159, I do believe you are correct about the reference to total profit, and that is an oversight on my end. Thanks.

It still makes me wonder though, what would happen if you closed each hour and rebuilt the grid. I've been around long enough to know I VERY rarely think of anything new in trading and forex. So I'm sure somebody has done this or seen it, and if it was a money maker, somebody here would be harping on about it. I'm still tempted to test it out, if for no other reason than to add to my education of what NOT to do -- and more importantly, why.
  • #134
  • Apr 3, 2008 1:59pm Apr 3, 2008 1:59pm
I wrote above, "Closed each hour and rebuilt the grid." I don't really mean that. Obviously, there are times the market is quiet and this just locks in loss after loss. More thought has to go into when to do it. Maybe something like, if you are at a loss but with less than 4 total positions open, then you keep them open... or maybe you close them if they are unbalanced... some kind of variation like that. --- Never a lack of things to look at around here.

... And of course, we shouldn't forget that the system looks like a long term winner just the way it is.
  • #135
  • Apr 5, 2008 9:28am Apr 5, 2008 9:28am
Wow, bad week for this system hu? Here are the results from the core version for April 2nd, sorry for the delay, was out of town.
Attached Image (click to enlarge)
Click to Enlarge Name: April 2.jpg Size: 301 KB
  • #136
  • Apr 5, 2008 9:39am Apr 5, 2008 9:39am
Remember, that these results are considering the core version rules in which you do not intervene in 24 hours. A great spike such as the one we experienced on April 3rd would give a great opportunity to profit from it if we applied the advanced techniques.
Attached Image (click to enlarge)
Click to Enlarge Name: April 3.jpg Size: 298 KB
  • #137
  • Apr 5, 2008 9:44am Apr 5, 2008 9:44am
April 4th -NO TRADE DUE TO NFP-

If you had had some positions open you would not have closed them yet. The NFP would provide a great opportunity to close these due to increased volatility and volume.

Better trades expected for next week. Will try to answer some of your post during the weekend.

Have a great weekend!
  • #138
  • Apr 5, 2008 2:42pm Apr 5, 2008 2:42pm
Isy,

I see you are saying that its a not trade during NFP. I am wondering what other no trades we would consider. I find that in todays world we have big economic news almost every day that will move this pair. For instance, wearlier in the wee there was big numbers out of Great Britain, that had a big impact, yet we were in the trade? Your previous results had the system trade every day from what I saw, so why now take out certain news days?
  • #139
  • Apr 5, 2008 6:13pm Apr 5, 2008 6:13pm
Quoting scottyb159
Isy,

I see you are saying that its a not trade during NFP. I am wondering what other no trades we would consider. I find that in todays world we have big economic news almost every day that will move this pair. For instance, wearlier in the wee there was big numbers out of Great Britain, that had a big impact, yet we were in the trade? Your previous results had the system trade every day from what I saw, so why now take out certain news days?
The Idea behind not trading NFP has more to do with spread widening than with volatility. Actually volatility is good, but spreads at oanda tend to widen anywhere from 10 to 100 pips (as claimed by some), it regularly widens 20 pips, so we do not want to see our orders being triggered not because of price action but due to spread widening. This backtest like any other is not 100% accurate, since we do not know many factors that were prevalent at the moment, so it's just an estimate of what could have happened during this period if a certain system had been traded, with an 80 to 95% accuracy. But, yes. If you would like to increase accuracy you would have to check all possible dates in which the spread widened and due to this factor you did not enter trade orders. But then also have an analysis of the trades that you would have gotten out from and the ones that you would have stayed till profit was met as the system requires it. You would have to analyze each day separately.

It is a good point though, and I will try to include in our forward testing the exits and advanced techniques at play as much in real time as possible.
  • #140
  • Apr 5, 2008 10:23pm Apr 5, 2008 10:23pm
i was thinking if we should adjust the lot sizes for orders further away from current prices to make it smaller.

so lot sizes could look something like this:
1st 0.15
2nd 0.13
3rd 0.11
4th 0.09
5th 0.07

As trends will slowly lose their strength, those at the extreme ends would not affect affect much if they were to be triggered and turn negative
0 traders viewing now