Best Exit Strategies for Forex market ?

[Barramundi]

imo peoples' plans should encompass all three strategies - entry, exit and re-entry.

And presumably a re-exit strategy?

 

[barjon]

And presumably a re-exit strategy?

i guess so, barra

 

[Fleming1]

Hi,One may attack after going a couple of steps back. I think this is the much risk managing step in Forex trading which i found in this thread.Good post.

 

[Jason101]

hey A

agreed - splitting 50/50 (or whatever % suits) is definitely the best of both worlds and can also be applied to entry methodology as well of course

yes its advanced stuff and not suitable for newbies but these kind of Discussions alongside MM are for me where the holy grail is for profitable trading vs finding the next great indicator/signal

N

Hi NVP,

I am a big fan of scaling out. But I guess here you are talking about scaling in.

This is something I have never really done (unless you count a new entry on an instrument I still have a trail on).

I know a lot of traders do scale in, it is something I would like to find out more about. Could you describe the methods you use for this please?

 

[Jason101]

As for the topic my scaling out exits are:

I have 5 lots and 2 targets into the longer term trend.
I exit 2 lots at 1:1, and bring the remaining 3 lots to b/e.

Then I exit 2 more lots at my 1:2 target and leave the last one as a runner, if a new entry sets up I will move this last (runner) stop to my new entry stop. (If longer term trend shows a sign of possible reversal I will exit this runner, without waiting for a new set up.)

(If my first 1:1 signal is generated at much less than 1:1, I will exit all at whatever the market has given, this for me is an indication the trend is now ebbing.)

 

[NVP]

Hi NVP,

I am a big fan of scaling out. But I guess here you are talking about scaling in.

This is something I have never really done (unless you count a new entry on an instrument I still have a trail on).

I know a lot of traders do scale in, it is something I would like to find out more about. Could you describe the methods you use for this please?

I dont scale in, but I suspect many do.......would make life very complicated though if yuo were also watching to scale out

hahahaha
N

 

[NVP]

As for the topic my scaling out exits are:

I have 5 lots and 2 targets into the longer term trend.
I exit 2 lots at 1:1, and bring the remaining 3 lots to b/e.

Then I exit 2 more lots at my 1:2 target and leave the last one as a runner, if a new entry sets up I will move this last (runner) stop to my new entry stop. (If longer term trend shows a sign of possible reversal I will exit this runner, without waiting for a new set up.)

(If my first 1:1 signal is generated at much less than 1:1, I will exit all at whatever the market has given, this for me is an indication the trend is now ebbing.)

excellent .....

this Is why I created this thread.......experienced traders have long moved on from the eternal question of "does my entry signal work ?" and into optimisation of Trades

N

 

[Adamus]

Hi NVP,

I am a big fan of scaling out. But I guess here you are talking about scaling in.

This is something I have never really done (unless you count a new entry on an instrument I still have a trail on).

I know a lot of traders do scale in, it is something I would like to find out more about. Could you describe the methods you use for this please?

I think the Turtles scaled in and pyramided their positions. They operated on a long time frame though with their long term trend following.

There have been discussions on T2W about increasing your position as a form of money management, but I'd say it's not something that is on-topic for this thread, unless you are discussing (like NVP mentions) whether you are still entering or whether you have started exiting.

 

[NVP]

yep the Turtles did

jees thats a long time ago now

N

 

[NVP]

actually some recent research and testing I have been doing reminded me of another factor I rate very highly in assessing Exit positions

Average True range of the Time period/pair(s) you are trading

if you are lucky enough to get extended (or whatever % of ATR you set) it may be a good idea to lock in some profits :smart:

N

 

[Adamus]

actually some recent research and testing I have been doing reminded me of another factor I rate very highly in assessing Exit positions

Average True range of the Time period/pair(s) you are trading

if you are lucky enough to get extended (or whatever % of ATR you set) it may be a good idea to lock in some profits :smart:

N

I find ATR can be great, simple and all you need but then at other times it just gets out of synch with itself and screws up every trade I try it on. So I start looking at it to see what's wrong - is the number of bars too low? 14 bars doesn't work but then 100 does. Or maybe instead of using the time frame of the bars I'm looking at, it seems more suitable to use a higher time frame with a bit shaved off.

I guess that's just the subconcious desire to be right all the time, driving me to complicate it all when I should be happy with what I've got and should KISS.

 

[NVP]

yep yep yep

the devil is in the detail......

N

 

[Joey25]

'Momentum' strategy: Exit after i days

'Counter-trend' strategy: Exit after j days

i > j

 

[Vorbis]

I know a lot of traders do scale in, it is something I would like to find out more about. Could you describe the methods you use for this please?

Check out FTI's thread "Technical Analysis Fallacy" over on FF, he's a bit of a specialist at scaling in and out.

Average True range of the Time period/pair(s) you are trading

This is my preferred method. I like 1XATR for SL and 2XATR and 3XATR for PT's. In my testing these performed as well as anything else except trailed Stops which entail longer draw-downs.

For very low drawdowns I found a 5-bar timed exit was ideal but not as profitable in the long run because you end up cutting off too many late-starters before they get chance to go.

 

[Joey25]

In the article "Beating the Foreign Exchange Market" (Sweeney, Journal of Finance 1986), filter rules are discussed, in this case as entry signals, but I reckon they're potentially good as exit signals.

The idea is simple - a signal is triggered by a fixed percentage move such as 0.5%, 1.0%, 2.0% etc. off a recent high or low.

The definition of "recent high or low" has various parameterisations, as does the percentage retrace.

 

[NVP]

Check out FTI's thread "Technical Analysis Fallacy" over on FF, he's a bit of a specialist at scaling in and out.



This is my preferred method. I like 1XATR for SL and 2XATR and 3XATR for PT's. In my testing these performed as well as anything else except trailed Stops which entail longer draw-downs.
For very low drawdowns I found a 5-bar timed exit was ideal but not as profitable in the long run because you end up cutting off too many late-starters before they get chance to go.

thats real nice V - and easy to follow...just out of interest have you done any work to establish what the 2X ATR equals say as a standard deviation of the ATR ?

I fancy something that say looks at the ATR and the T1, T2 etc related to the Standard deviation - eg 1Xsd , 2Xsd and 3xsd for those big tail trades

just a thought..I am not a methematician

N

 

[NVP]

In the article "Beating the Foreign Exchange Market" (Sweeney, Journal of Finance 1986), filter rules are discussed, in this case as entry signals, but I reckon they're potentially good as exit signals.

The idea is simple - a signal is triggered by a fixed percentage move such as 0.5%, 1.0%, 2.0% etc. off a recent high or low.

The definition of "recent high or low" has various parameterisations, as does the percentage retrace.

yep.........% could be placed in the same school as ATR as they both look at the historical performance of the instrument to indicate the future volatility

N

 

[Vorbis]

thats real nice V - and easy to follow...just out of interest have you done any work to establish what the 2X ATR equals say as a standard deviation of the ATR ?

Not looked at standard deviation, prefer to keep things simple. One thing I discovered is that it doesn't much matter what you use because when you adjust, you end up swapping Hit Rate for RR Ratio. the exception was 1XR which gave very poor results (trend bias not being utilised?) and with 4XR+ there is a slight drop off of profitability coupled with the inevitably greater DD.

The main thing I think is to pick something that works reasonably well and apply it consistently. Correctly identifying the direction of flow is far more important IMO; tinkering with the PT doesn't seem to make as much difference as you'd think, so long as it's not outrageous.

 

[Joey25]

yep.........% could be placed in the same school as ATR as they both look at the historical performance of the instrument to indicate the future volatility

N

To clarify, the filter rule exit is as follows:

Example Long

(The defintion of a "local maximum" I use here is a price surrounded by lower prices to the left and to the right).

Buy at 100 on day 1
Day 2: price 101
Day 3: price 102
Day 4: price 101 ------- at this point a local maximum has been formed at 102

The filter rule exit is then a % fall from 102, so if we use 5% it needs to fall from 102 to 96.9 before the exit signal is triggered.

If a new local maximum is formed that is higher than 102 before 96.9 is hit then a 5% fall from this new higher-high is required.

 

[NVP]

Not looked at standard deviation, prefer to keep things simple. One thing I discovered is that it doesn't much matter what you use because when you adjust, you end up swapping Hit Rate for RR Ratio. the exception was 1XR which gave very poor results (trend bias not being utilised?) and with 4XR+ there is a slight drop off of profitability coupled with the inevitably greater DD.

The main thing I think is to pick something that works reasonably well and apply it consistently. Correctly identifying the direction of flow is far more important IMO; tinkering with the PT doesn't seem to make as much difference as you'd think, so long as it's not outrageous.

agreed - as with entries if you keep changing the rules you create more problems

simple and mechanical......warts an all

N