Automated Grid Trading | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Flexy Grid EA (Grid Trading) 1,363 replies

Grid - New indicator that replaces chart grid 60 replies

Automated or semi automated trading? 3 replies

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

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

Attachments: Automated Grid Trading
Exit Attachments

Automated Grid Trading

  • #1
  • First Post: Mar 10, 2007 2:59pm Mar 10, 2007 2:59pm
As I am working on the automated grid trading for 1 year, I want to open discussion about it!

Does anyone have any experience about this topic?

I give u a brief description of it:

  1. Use very small lotsize (nano trades = 0.01 lot size, 1000$ for standard, 100$ for mini accounts). This system runs tick based, not candle based. It opens maximum 1 trade per tick. All trades are opened at market, no stoploss, no limitorders.
  2. Start with 1 buy at market and 1 sell at market. They only differ with a currency spread.
  3. From that point, no other buys will be placed below first buy, no other sell above first sell.
  4. Recalculate profit for each tick, it there is negative profit, then place a trade at market (buy if above or sell if below) BUT only if there is still vacant pip position.
  5. You then have a group of buys on the upper part and a group of sells on the lower part. If there is finally positive profit, it stops with trading and leaves a group of trades. As long as the market moves away from them, the profit becomes larger.

But there are also risks with this trading method:

  1. When market changes direction, going back into grid, the profit becomes back negative: it continues to open trades on another vacant pip positions, making a grid larger until the profit finally becomes again positive.
  2. The number of trades can be very large, that the available margin could become zero.
  3. It can take hours, days or weeks time before the total profit target should be reached.
  4. Too many trades with too much margin can blow out accounts. So, it’s important to have big leverage (400:1 or 200:1).
  5. Also important: IT MUST HAVE FAST CONNECTION WITH THE FX SERVER! Because this system uses only trades at market.
  6. Some brokers allow only a LIMITED number of open trades.

In order to try to prevent these risks, I improved this system with starting conditions:

  1. Start immediately
  2. Start if ask has reached VALUE1
  3. Start if bid has reached VALUE2
  4. Start if ONE of BOTH condition is true

The VALUES are to be defined where it’s best to start from, after studying charts.

That’s all where I am currently developing a stand-alone robot with this system. Also I have backtested this system using MT4 terminal. I use both MT3 and MT4 demo accounts.

Any questions and suggestions are welcome!

Regards,
Jeff

For hiring me to program EA's, please PM me.
  • #2
  • Mar 10, 2007 3:49pm Mar 10, 2007 3:49pm
it seems to backtest this idea you'd need tick data, which can be hard to come by.
  • #3
  • Mar 10, 2007 3:55pm Mar 10, 2007 3:55pm
rnhturner posted another grid strategy from other forums. You can take
a look at that also.

Originally Posted by rnhturner
1. 10 buys 11 sells = 1 sell in profit is no where near what I am talking about. This is not how I am doing this. I am saying as price goes up 100 pips without retracement or closing trades you will have 20 buys AND 20 sells!! At this very moment the only ones in profit are your buys. I am saying you can manage your grid at first by:

a. closing the highest profitable buy and closing the lowest losing trade that is outside of the brokers new order zone. (meaning we will eventually have to replace this position with a new order and to be able to do this with an EA we have to assume that the broker will not let us plcea new order unless it is outside of 5 pips) Theoretically this would be closing the buy that is 100 pips up and the sell that is only 5 pips negative. So without doing anything else we have an account balance (i know not equity) plus 95 pips. SO FAR THIS IS A SERO SUM GAME (outside of the cost of the spread)

b. now we let the makegrid do its things and it will place a new buy order at the same place where it was originally bought. (so a buy placed at 1.3000 that was closed at 1.3100 will place a new buy order at 1.3000) This fills in that gap of the buy. The same thing is done with the sell.


c. THIS IS WHAT I NEED HELP WITH PROGRAMMING WISE: If I can't find the help here I will go else where to find it. I need the EA to place both a new buy order at 1.3000 and a sell at 1.3000 so when/if price ever gets back to that point the grid stays hedged. If price bounces in this area back and forth like forex does we make our equity grow! Because it is no longer a zero sum game! If and when price reaches all the way back to the 1.3000 level we will now have one buy and 2 sells at that level, but at the same time we are completely hedged with even positons on both sides of the price. As long as price ranges we are able to caputure profit at these postions over and over and over.

d. Now instead of thinking of only one position doing this, think of all the other levels that we could have taken profit on. at that same point where we closed the buy at 100 pips profit we had another buy that was at 95 pips profit and a sell at -10 pips.

It might very well looked like this below:

100 - 5 = 95
95 -10 = 85
90 -15 = 75
85-20 = 65
80 - 25 = 55
75- 30 = 45
70 - 35 = 35
65- 40 = 25
60 - 45 = 15
55 - 50 = 5

Total closed trades at this one level = 500 pips minus the spread cost of 2 pips per trade = -40 pips = Total closed trade after expenses = 460 pips.

BUT we also have new orders at everyone of these positions so if price retraces all the way down to 1.3000 we will be able to capture about 460 more pips. If price goes back up we do it again! Eventually these postions equal positive equity.

If price continues to go below or above the examples above we are basically only paying the spread cost to get into these positions and of course margin cost if not using fxdd.


2. This is not as slow as a process as you might think. You can actually close postions everyday! When price ranges you make money. When price trends you will still be able to close trades everyday!


I only bolded the numbers and leters to some how make my cluttered resonse a litte less cluttered. Not sure it helped!
  • #4
  • Mar 11, 2007 2:28pm Mar 11, 2007 2:28pm
Quoting aicccia
it seems to backtest this idea you'd need tick data, which can be hard to come by.
If you regularly backtest it for months, the MT4 terminal use old tick data in your PC and fill them with new tick data from the server. I use additional EA with SAR indicator for backtesting this grid trading system: if a long SAR of at least 10 points has been broken, it has a better point than a random one to start trading. If a total profit has reached a defined limit, then all trades are closed and then waiting for the good point to start next trading, etc.

Here I have some backtesting results for 4 months, it's the highest risk settings so that the account has just not got blown away. I used 0.01 standard lotsize per 2500$ equity. The profit limit is 10000 * 0.01 standard lotsize per 2500$ equity. Example: with 30000$ deposit the profit limit = 10000 * 0.01 * (30000/2500) = 1200$


Strategy Tester Report: NanoPips


Symbol: GBPUSD
Period: 1 Hour (H1) 2006.09.01 00:00 - 2007.02.09 20:00 (2006.09.01 - 2007.02.10)
Model Every tick (based on all available least timeframes with fractal interpolation of every tick)

Bars in test: 3683
Ticks modelled: 805250
Modelling quality:85.69%

Initial deposit: 30000.00

Total net profit: 146747.31
Gross profit: 731121.68
Gross loss: -584374.37
Profit factor: 1.25
Expected payoff: 34.66

Absolute drawdown: 12083.48
Maximal drawdown: 76637.82 (50.95%)
Relative drawdown: 52.51% (75900.89)

Total trades: 4234
Short positions (won %): 2080 (70.53%)
Long positions (won %): 2154 (76.14%)

Profit trades (% of total): 3107 (73.38%)
Loss trades (% of total): 1127 (26.62%)
Largest profit trade: 997.68
loss trade: -1418.76
Average profit trade: 235.31
loss trade: -518.52
Maximum consecutive wins (profit in money): 162 (81786.79)
consecutive losses (loss in money): 63 (-75900.89)
Maximal consecutive profit (count of wins): 83102.40 (159)
consecutive loss (count of losses): -76637.82 (62)
Average consecutive wins: 71
consecutive losses: 26
Attached Image
For hiring me to program EA's, please PM me.
  • #5
  • Mar 11, 2007 3:03pm Mar 11, 2007 3:03pm
Quoting meltri
rnhturner posted another grid strategy from other forums. You can take
a look at that also.

Originally Posted by rnhturner
1. 10 buys 11 sells = 1 sell in profit is no where near what I am talking about. This is not how I am doing this. I am saying as price goes up 100 pips without retracement or closing trades you will have 20 buys AND 20 sells!! At this very moment the only ones in profit are your buys. I am saying you can manage your grid at first by:

a. closing the highest profitable buy and closing the lowest losing trade that is outside of the brokers new order zone. (meaning we will eventually have to replace this position with a new order and to be able to do this with an EA we have to assume that the broker will not let us plcea new order unless it is outside of 5 pips) Theoretically this would be closing the buy that is 100 pips up and the sell that is only 5 pips negative. So without doing anything else we have an account balance (i know not equity) plus 95 pips. SO FAR THIS IS A SERO SUM GAME (outside of the cost of the spread)

b. now we let the makegrid do its things and it will place a new buy order at the same place where it was originally bought. (so a buy placed at 1.3000 that was closed at 1.3100 will place a new buy order at 1.3000) This fills in that gap of the buy. The same thing is done with the sell.


c. THIS IS WHAT I NEED HELP WITH PROGRAMMING WISE: If I can't find the help here I will go else where to find it. I need the EA to place both a new buy order at 1.3000 and a sell at 1.3000 so when/if price ever gets back to that point the grid stays hedged. If price bounces in this area back and forth like forex does we make our equity grow! Because it is no longer a zero sum game! If and when price reaches all the way back to the 1.3000 level we will now have one buy and 2 sells at that level, but at the same time we are completely hedged with even positons on both sides of the price. As long as price ranges we are able to caputure profit at these postions over and over and over.

d. Now instead of thinking of only one position doing this, think of all the other levels that we could have taken profit on. at that same point where we closed the buy at 100 pips profit we had another buy that was at 95 pips profit and a sell at -10 pips.

It might very well looked like this below:

100 - 5 = 95
95 -10 = 85
90 -15 = 75
85-20 = 65
80 - 25 = 55
75- 30 = 45
70 - 35 = 35
65- 40 = 25
60 - 45 = 15
55 - 50 = 5

Total closed trades at this one level = 500 pips minus the spread cost of 2 pips per trade = -40 pips = Total closed trade after expenses = 460 pips.

BUT we also have new orders at everyone of these positions so if price retraces all the way down to 1.3000 we will be able to capture about 460 more pips. If price goes back up we do it again! Eventually these postions equal positive equity.

If price continues to go below or above the examples above we are basically only paying the spread cost to get into these positions and of course margin cost if not using fxdd.


2. This is not as slow as a process as you might think. You can actually close postions everyday! When price ranges you make money. When price trends you will still be able to close trades everyday!


I only bolded the numbers and leters to some how make my cluttered resonse a litte less cluttered. Not sure it helped!
Great idea, if I understand this trading system, I should put a buy and a sell each 5 pips position and if a market moves away more than 100 pips going back only a part (new range after trending), I should do the following:

  1. I suppose the new market range will be 100 pips wide, then I should close 50% of big profiting buys and 50% of small losing sells, all inside new market range. And putting new orders on the same positions.
  2. All other pairs (buy - sell) that are outside new market range are to be closed with spread losses.

But why don't we put smaller trades on EACH pip position? Then you should try to have more -or better all- buys and sells on the same positions without spread losses. You can close a buy-sell pair at the same time using OrderCloseBy command in MQL language. It's safer than to close them separately. Only they must have the same volume.

And thinking about EA, it will be a more complexer one than that I have written a simple grid trading system where I am telling about. My grid trading has no indicators at all, and this grid idea maybe needs a complex EA to decide where the market should range... and also an anti-spread EA.

Jeff

For hiring me to program EA's, please PM me.
  • #6
  • Mar 12, 2007 12:13am Mar 12, 2007 12:13am
Very intersting but it looks like that your griid is only 100 pips wide. what happens when the price moves out of this range?

Also can you tell me in what forum can I find rnhturner? I would really want to talk to him about his strategy, if you cant post the link here can you just PM me for it.

Thanks,
Mikhail

Quoting meltri
rnhturner posted another grid strategy from other forums. You can take
a look at that also.

Originally Posted by rnhturner
1. 10 buys 11 sells = 1 sell in profit is no where near what I am talking about. This is not how I am doing this. I am saying as price goes up 100 pips without retracement or closing trades you will have 20 buys AND 20 sells!! At this very moment the only ones in profit are your buys. I am saying you can manage your grid at first by:

a. closing the highest profitable buy and closing the lowest losing trade that is outside of the brokers new order zone. (meaning we will eventually have to replace this position with a new order and to be able to do this with an EA we have to assume that the broker will not let us plcea new order unless it is outside of 5 pips) Theoretically this would be closing the buy that is 100 pips up and the sell that is only 5 pips negative. So without doing anything else we have an account balance (i know not equity) plus 95 pips. SO FAR THIS IS A SERO SUM GAME (outside of the cost of the spread)

b. now we let the makegrid do its things and it will place a new buy order at the same place where it was originally bought. (so a buy placed at 1.3000 that was closed at 1.3100 will place a new buy order at 1.3000) This fills in that gap of the buy. The same thing is done with the sell.


c. THIS IS WHAT I NEED HELP WITH PROGRAMMING WISE: If I can't find the help here I will go else where to find it. I need the EA to place both a new buy order at 1.3000 and a sell at 1.3000 so when/if price ever gets back to that point the grid stays hedged. If price bounces in this area back and forth like forex does we make our equity grow! Because it is no longer a zero sum game! If and when price reaches all the way back to the 1.3000 level we will now have one buy and 2 sells at that level, but at the same time we are completely hedged with even positons on both sides of the price. As long as price ranges we are able to caputure profit at these postions over and over and over.

d. Now instead of thinking of only one position doing this, think of all the other levels that we could have taken profit on. at that same point where we closed the buy at 100 pips profit we had another buy that was at 95 pips profit and a sell at -10 pips.

It might very well looked like this below:

100 - 5 = 95
95 -10 = 85
90 -15 = 75
85-20 = 65
80 - 25 = 55
75- 30 = 45
70 - 35 = 35
65- 40 = 25
60 - 45 = 15
55 - 50 = 5

Total closed trades at this one level = 500 pips minus the spread cost of 2 pips per trade = -40 pips = Total closed trade after expenses = 460 pips.

BUT we also have new orders at everyone of these positions so if price retraces all the way down to 1.3000 we will be able to capture about 460 more pips. If price goes back up we do it again! Eventually these postions equal positive equity.

If price continues to go below or above the examples above we are basically only paying the spread cost to get into these positions and of course margin cost if not using fxdd.


2. This is not as slow as a process as you might think. You can actually close postions everyday! When price ranges you make money. When price trends you will still be able to close trades everyday!


I only bolded the numbers and leters to some how make my cluttered resonse a litte less cluttered. Not sure it helped!
  • #7
  • Mar 12, 2007 4:32am Mar 12, 2007 4:32am
Guys, this will not work. You just keep adding positions. You close some with profit but your floating loss always rises. You loose the spread on all those orders.

Be very careful, always try new things on demo first...
  • #8
  • Mar 12, 2007 4:41am Mar 12, 2007 4:41am
Quoting mikhaildgreat
Very intersting but it looks like that your griid is only 100 pips wide. what happens when the price moves out of this range?

Also can you tell me in what forum can I find rnhturner? I would really want to talk to him about his strategy, if you cant post the link here can you just PM me for it.

Thanks,
Mikhail

Hello Mikhail,

just have a look at http://www.strategybuilderfx.com/sho...15936&page=106

There you can find rnhturner. I would really like seeing you building a working EA for this strategy.

bye

Michael
  • #9
  • Mar 12, 2007 5:17am Mar 12, 2007 5:17am
Quoting Docsniper
Hello Mikhail,

just have a look at http://www.strategybuilderfx.com/sho...15936&page=106

There you can find rnhturner. I would really like seeing you building a working EA for this strategy.

bye

Michael
Hi,

Nice to see you in this forum. How are you doing? are still using the Docsniper EA you had me customize for you? You would have earned quite a handsome profit carry trading with that EA for the past months.

Anyways with the above strategy it sound good but im not too clear on how it works though... the grid is only 100 pips wide... what do you do when the price moves away from your grid? do you wait for it to come back?

Mikhail
  • #10
  • Mar 12, 2007 5:30am Mar 12, 2007 5:30am
The scientist is right. I had studied grids for a very long time, and they will not work as stand alone applications. TDION had this clever way of putting it, "a mathematical trap on the price"... this is the theory to what grids attempt to achieve. However, and I've stated this before... the markets are not a 2+2=4 type of environment, the grid is. Please don't misunderstand me - if a trader opens a 100K account, he can set up a grid and trade mini (or nano) lots and generate a less than attractive profit. Keeping in mind he will still be susceptible to long periods of draw down (some times months and possibly years). Come to think of it, he might have better luck (and bigger profits) with a martingale EA with that size of an account. If you guys are curious and you'd like to get into some 'brain aerobics', then you're definitely barking up the right tree. If you're trying to be 'slick' in an attempt to eliminate the risk of trading... think twice before you piss away much time and energy.

I will say this about having spent such a long time observing grids - great educational tool for cost averaging. Good luck, although I don't think luck will get the chance to be a factor in this little endeavor.

P.S. please don't reply to this message with comments like, "why are you being such a negative jerk!!!", if you do I will give you a virtual bitch slaphttp://www.forexfactory.com/images/icons/icon10.gif
I'm posting this as a benefit to the thread, I'm hoping you see it as such.

Peace out peeps,
MM
  • #11
  • Mar 12, 2007 7:23am Mar 12, 2007 7:23am
Quoting mikhaildgreat
Hi,

Nice to see you in this forum. How are you doing? are still using the Docsniper EA you had me customize for you? You would have earned quite a handsome profit carry trading with that EA for the past months.

Anyways with the above strategy it sound good but im not too clear on how it works though... the grid is only 100 pips wide... what do you do when the price moves away from your grid? do you wait for it to come back?

Mikhail

Thx man, I´m doing great. The EA works flawlessly and I managed to get out of the trades before the big downfall in the Yen-crosses.


I think the strategy of rnhturner is not limited to a 100 pip wide grid. He only used the 100 pip as example. As far as I understand, the grid he builds has no limitations concerning the range it covers. As you would be nearly always hedged, the only limitations would be spread cost.

bye

Michael
  • #12
  • Mar 12, 2007 7:52am Mar 12, 2007 7:52am
Doc I'm happy to hear you are doing good. I wish I also used your EA a few months back do you think it would still work? looks like after the JPY rate announcement carry trading has resumed...

With regards to this particular grid strategy I will have to study it more, I still can't seem to grasp the whole strategy yet.

Anyways good luck at the market,
Mikhail
  • #13
  • Mar 12, 2007 9:26am Mar 12, 2007 9:26am
Quoting mikhaildgreat
Doc I'm happy to hear you are doing good. I wish I also used your EA a few months back do you think it would still work? looks like after the JPY rate announcement carry trading has resumed...

With regards to this particular grid strategy I will have to study it more, I still can't seem to grasp the whole strategy yet.

Anyways good luck at the market,
Mikhail

I think it will work again, but I would start the carry trade as soon as the GBPJPY has dropped again significantly. Maybe around the end of the month (end of japanese fiscal year) we´ll see the 220-218. So, then I´ll start again this strategy with averaging in by daily buying one mini-lot.

Ok, sorry for offtopic, now concerning the topic of this thread:

Have a look at the thread at strategybuilder-fx-forum. Try to do the strategy manually, it seems to work (at least for me, but three weeks of forward testing are not enough to judge the concept of this grid).
  • #14
  • Apr 14, 2007 3:06pm Apr 14, 2007 3:06pm
What currencies work best with grid systems?
24 hours in a day – 24 bottles of beer in a case – Coincidence? I think not
  • #15
  • Last Post: Nov 3, 2012 9:37am Nov 3, 2012 9:37am
Hi fxtrader625,

Have you seen this grid system, the idea sounds nice, also abstract. What's your opinion? www .youtube.com/watch?feature=player_embedded&v=jAgZ-NTM1a8
0 traders viewing now