I've read bunch of threads on the Grid Trading EA. I decided to put my thoughts together and hopefully someone can either confirm or correct my assumptions.
The Grid System in a Nut Shell:
To start, pick a random currency pair that I like, say GBP/JPY (don't ask me why, but i like this pair for some reason). GBP/JPY has about 6 pips spread, and it can get larger or smaller depends on the volitility and liquidity.
I use 20 pips spacing for the grid, this might be a little too narrow for manual trading, but once I finish the EA, it should be okay.
Say the current price is at 210.78.
you put 5 buy stops @210.90, 211.10, 211.30, 211.50, and 211.70
you put 5 sell stops @210.70, 210.50, 210.30, 210.10, and 209.90
I have no SL and no visible TP (i usually close it manually when it hits about 25 pip TP)
Reset the Grid:
the basic setup is simple, i think the discussion and strategies comes in when you have to decide when to close some dangling orders and when and where to put in new buy and sell stops.
here is the strategy that i use:
if a grid is taken out up or down, and the price is more than the grid that has been taken up, then put in set stop order in the opposite direction. (e.g. if buy stop at 210.90 is taken out, and the price is at 211.10 or above, then put in a new sell stop at 210.90...reason being most brokers require 10 pips difference in putting in stops from the bid)
in weird senarios (but do happen from time to time), you have a dangling position, (e.g. 1 long at 211.10 and 1 short at 210.50, but the price is currently at 210.45...this scenario does happen more often than you think), i would still put 5 buy stops 20 pips apart, but i will skip 211.10 because i already have a dangling position)
here is an example:
let's say the price went up 20 pips, now at 210.98, so one stop buy hits, and now you net 1 long position @210.90. i wouldn't do anything until the following happens:
1) if the price keep going up, say it hits 211.18. now i'll have 2 long positions, at 211.10 and 210.90. at this point i would close the 210.90 at a profit (manually around 19-20 pips profit after taken out the spread and commission). and i would also add another sell stop @ 210.90. so now i am long only 1 open position.
2) if the price did not go up, it went down to 210.60. at this point, the 210.70 sell stop would hit, i would be at net neutral position with locked in 20 pips loss (long at 210.90 and short at 210.70). at this point i'd do nothing.
it is also possible that long stop 211.10 hit and you didn't close the 210.90 at a profit, so now you're net long 1 poition (long 211.10, 210.90 and short 210.70). this is okay because if the market moves against you by going down, there is a sell stop at 210.50. i wouldn't worry about this situation either.
This setup works great is it is a ranging market, but i think it would be okay in a trending market. In a trending market, I just have to be sure when i close sell orders with profit, i would close it so that i would be net neutral. say, the price has moved down 50 pips, with 1 net long at a huge loss and 2 net shorts with profit. i would only close 1 short position at profit, or wait til the price hit the next sell stop grid. as long as i keep it net neutral, i am only losing small amount of money in swaps.
I think this is the gist of the system...if i am missing anything, please feel free to add.
Possible Problems with the system:
i am compling a lsit of possible scenarios which is likely to break down the grid system.
1) news events. unless you have grid spacing about 150pips, you're gonna get killed when the spread opens up around news time. the only solution that i can think of is to make sure you're net neutral before the news event and close all your pending orders and put them back in when the spread is back to normal.
2) weekend gap. market can gap up or down on sunday open...i guess you can use the same strategy as news events for the weekend gap.
3) increasing hedged positions. i think this is the big problem with grid trading. you get 1 extra hedged position whenever the price just touch the grid and reverse. if you're in a strong trend with some pullbacks, you'll end up with a lot of hedged positions. this is a big problem because even if you have hedged positions of the same currency, most brokers will still charge you swaps (positive interest will always be a little smaller than the negative interest...this is just the rule of the game). this is the major gripe i have with the grid trading. imagine you start grid trading a year a go, and the price just goes in one direction, you'll probably end up with 20-30 open hedged position. how do you solve this problem?
4) unexpected events. like news events, but unexpected...could be bear stearns announcement or terrorist bombing...either way when this happens, the spread and the volitility will destroy your grid, and you have no time to make your position net neutral or take out the pending orders. i guess when this happens, you're just fucked. if you're doing grid trading, just hope nothing major happens in weird hours.
Questions and Discussions:
i've outline the grid system in a nutshell, proposed couple of trading examples, and problematic scenarios of this system. what i like to know is how some of your strategies to unwind hedged positions. would you close the position anyways or you just leave them all open and call it "critical mass" or "core grid positions". eventually i think you'll get nickle and dimed by the swaps differential.
will the system work on any major pair? or you only choose ranging pairs like eur/gbp?
The Grid System in a Nut Shell:
To start, pick a random currency pair that I like, say GBP/JPY (don't ask me why, but i like this pair for some reason). GBP/JPY has about 6 pips spread, and it can get larger or smaller depends on the volitility and liquidity.
I use 20 pips spacing for the grid, this might be a little too narrow for manual trading, but once I finish the EA, it should be okay.
Say the current price is at 210.78.
you put 5 buy stops @210.90, 211.10, 211.30, 211.50, and 211.70
you put 5 sell stops @210.70, 210.50, 210.30, 210.10, and 209.90
I have no SL and no visible TP (i usually close it manually when it hits about 25 pip TP)
Reset the Grid:
the basic setup is simple, i think the discussion and strategies comes in when you have to decide when to close some dangling orders and when and where to put in new buy and sell stops.
here is the strategy that i use:
if a grid is taken out up or down, and the price is more than the grid that has been taken up, then put in set stop order in the opposite direction. (e.g. if buy stop at 210.90 is taken out, and the price is at 211.10 or above, then put in a new sell stop at 210.90...reason being most brokers require 10 pips difference in putting in stops from the bid)
in weird senarios (but do happen from time to time), you have a dangling position, (e.g. 1 long at 211.10 and 1 short at 210.50, but the price is currently at 210.45...this scenario does happen more often than you think), i would still put 5 buy stops 20 pips apart, but i will skip 211.10 because i already have a dangling position)
here is an example:
let's say the price went up 20 pips, now at 210.98, so one stop buy hits, and now you net 1 long position @210.90. i wouldn't do anything until the following happens:
1) if the price keep going up, say it hits 211.18. now i'll have 2 long positions, at 211.10 and 210.90. at this point i would close the 210.90 at a profit (manually around 19-20 pips profit after taken out the spread and commission). and i would also add another sell stop @ 210.90. so now i am long only 1 open position.
2) if the price did not go up, it went down to 210.60. at this point, the 210.70 sell stop would hit, i would be at net neutral position with locked in 20 pips loss (long at 210.90 and short at 210.70). at this point i'd do nothing.
it is also possible that long stop 211.10 hit and you didn't close the 210.90 at a profit, so now you're net long 1 poition (long 211.10, 210.90 and short 210.70). this is okay because if the market moves against you by going down, there is a sell stop at 210.50. i wouldn't worry about this situation either.
This setup works great is it is a ranging market, but i think it would be okay in a trending market. In a trending market, I just have to be sure when i close sell orders with profit, i would close it so that i would be net neutral. say, the price has moved down 50 pips, with 1 net long at a huge loss and 2 net shorts with profit. i would only close 1 short position at profit, or wait til the price hit the next sell stop grid. as long as i keep it net neutral, i am only losing small amount of money in swaps.
I think this is the gist of the system...if i am missing anything, please feel free to add.
Possible Problems with the system:
i am compling a lsit of possible scenarios which is likely to break down the grid system.
1) news events. unless you have grid spacing about 150pips, you're gonna get killed when the spread opens up around news time. the only solution that i can think of is to make sure you're net neutral before the news event and close all your pending orders and put them back in when the spread is back to normal.
2) weekend gap. market can gap up or down on sunday open...i guess you can use the same strategy as news events for the weekend gap.
3) increasing hedged positions. i think this is the big problem with grid trading. you get 1 extra hedged position whenever the price just touch the grid and reverse. if you're in a strong trend with some pullbacks, you'll end up with a lot of hedged positions. this is a big problem because even if you have hedged positions of the same currency, most brokers will still charge you swaps (positive interest will always be a little smaller than the negative interest...this is just the rule of the game). this is the major gripe i have with the grid trading. imagine you start grid trading a year a go, and the price just goes in one direction, you'll probably end up with 20-30 open hedged position. how do you solve this problem?
4) unexpected events. like news events, but unexpected...could be bear stearns announcement or terrorist bombing...either way when this happens, the spread and the volitility will destroy your grid, and you have no time to make your position net neutral or take out the pending orders. i guess when this happens, you're just fucked. if you're doing grid trading, just hope nothing major happens in weird hours.
Questions and Discussions:
i've outline the grid system in a nutshell, proposed couple of trading examples, and problematic scenarios of this system. what i like to know is how some of your strategies to unwind hedged positions. would you close the position anyways or you just leave them all open and call it "critical mass" or "core grid positions". eventually i think you'll get nickle and dimed by the swaps differential.
will the system work on any major pair? or you only choose ranging pairs like eur/gbp?
