The Best Way To Choose Exits | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Broker Comparison Table/chart - best way to choose the right broker 14 replies

The Best Way To Choose Lot Size 15 replies

Why Use Multiple Exits? by Chuck LeBeau 0 replies

Importance of Exits, by Chuck LeBeau 0 replies

Stops Exits And Trade Management 3 replies

The Best Way To Choose Exits

  • #1
  • First Post: Oct 11, 2006 5:39pm Oct 11, 2006 5:39pm
Ok I feel that there was some good feedback regarding "The Best Way To Choose Lot Sizes", so I'd thought I would post another thread, this time regarding exits.

It seems to me a great irony of trading that we spend hours studying and studying, until we have those perfect conditions as given by our indicators to enter. So we place an order when (for example) RSI crosses 50, 5EMA crosses 10EMA, and my grandma falls asleep. When do we usually exit. When the price goes up 50 or goes down 25. How much sense does this make having conditions for our entries but random numbers for our exits?

I think that we should base our exits on the same indicators we base our entries on. Why would we stop monitoring the indicators after we entered the trade and blindly wait for the trade to either close in the black or red?

So my answer to this problem is to develop levels or conditions for our exits as well as our entries. I believe that one can increase winning percentage and possibly decrease the average loss. I am not yet sure what these levels are or how to determine them, but I believe this philosophy will be more profitable than the widely used current method.

I appreciate any and all suggestions,
andersenws

Note: In this thread and "The Best Way To Choose Lot Sizes", I am speaking in terms of mechanical trading systems and EAs. I realize many, perhaps all, successful traders use discretion in their trading. However, I am currently studying mechanical systems, and am not concerned with discretion right now.
"Youth is the trustee of prosperity." - Benjamin Disraeli
  • #2
  • Oct 11, 2006 5:57pm Oct 11, 2006 5:57pm
Quoting andersenws
I think that we should base our exits on the same indicators we base our entries on. Why would we stop monitoring the indicators after we entered the trade and blindly wait for the trade to either close in the black or red?
I couldn't agree more! Afterall, isn't an exit in one direction a potential entry in the opposite direction? I realize this may not always be the case but the fewer indicators you use and the more ways you find ways to use them in unique ways the better off you are.
  • #3
  • Oct 11, 2006 6:28pm Oct 11, 2006 6:28pm
Quoting andersenws
When the price goes up 50 or goes down 25. How much sense does this make having conditions for our entries but random numbers for our exits?

I will agree with you that many new traders (hell, maybe old traders too) pull targets and stop loss numbers right out of their ass. However, fixed S/L and Targets can be tested and statistics can be generated for them just like any other exit methodology.

You may indeed find that using the same indicator(s) and an opposing signal is the best exit. There are a lot of ways to trade, however, and my experience has been that some methods can do very well (even better) with fixed targets as compared to indicator exits.

I will say that fixed targets tend to do better with a breakeven floor amount and some type of trailing stop.
Secret Identity: Bill Sullivan
  • #4
  • Oct 11, 2006 10:33pm Oct 11, 2006 10:33pm
I am still leaning towards my original opinion, with this logic. If you rely on an indicator to tell you when a move is coming, doesn't it also make sense to rely on it to tell you when the move is over.

If we flipped around the current method (indicator entries, random exits), it's easy to see why some could call it ridiculous.

Buy EURUSD at 1.24xx and sell when RSI crosses below 50.

I would think that any trader would find this proposition ridiculous, yet this is what many of them do, only on the exit side rather than the entry side.

Also, I do not necessarily believe that you can forecast the magnitude of a move (which is basically what a TP is). You know a move is coming because of your indicator, so you should ride it out until your indicator signals the end or a reversal.

Still open to others' thoughts,
andersenws
"Youth is the trustee of prosperity." - Benjamin Disraeli
  • #5
  • Oct 11, 2006 10:39pm Oct 11, 2006 10:39pm
Quoting andersenws
I am still leaning towards my original opinion, with this logic. If you rely on an indicator to tell you when a move is coming, doesn't it also make sense to rely on it to tell you when the move is over.

If we flipped around the current method (indicator entries, random exits), it's easy to see why some could call it ridiculous.

Buy EURUSD at 1.24xx and sell when RSI crosses below 50.

I would think that any trader would find this proposition ridiculous, yet this is what many of them do, only on the exit side rather than the entry side.

Also, I do not necessarily believe that you can forecast the magnitude of a move (which is basically what a TP is). You know a move is coming because of your indicator, so you should ride it out until your indicator signals the end or a reversal.

Still open to others' thoughts,
andersenws

You are assuming an awful lot here. You know a move is coming? You know there is a certain probability of a move based on your backtesting. Why can't you know that a certain percentage of those moves go x number of pips?
Secret Identity: Bill Sullivan
  • #6
  • Oct 11, 2006 11:23pm Oct 11, 2006 11:23pm
My 2 pips on the topic of exit picking...

In Tom DeMark's book "The New Science of Technical Analysis" he talks about some of his favorite moving averages. For my example/thoughts i'll take two of his favorites, the 13 and 55. Now he dosn't use them in the typical "Cross" to enter a trade, rather he takes notice the first day both change directions in unison. Anyway, we are talking exit picking here, not entries, so i ramble. For a simple excersize (though this will take some quality manual backtesting/study time), take these same two MAs and take note of their location in time and space where they cross. Now instead of using Vega's moving Median MA with Fibo-Numbered-Extensions, take those same FNExt from the point of the cross and see if there is a correlation between the numbers and stalls/consolidations/reversals. Maybe you begin to see some patterns taking place on a 30m or 60m TF with these MAs? Do some more research, see some more correlations?? In the long run there are patterns that will be recognized by using the above method and one can base exit strategies/rules by these studies that will result in greater gains than simply picking past S/R or XX number of profit-pips.

This is how i choose my exits, and it is profitable rather than the moving fibo TP areas off a median MA.

Here is a quick post script insert of an EA based trade i've been experimenting with that did what i discussed above. You can see that the TP levels were set based off the MA intersection, then at that moment in time, the levels were calculated and TPs set in stone more or less, and the trade was then let run it's course taking profit at the Fibo numbered extensions. This is in contrast to simply visualizing where the currency is in reference to the moving median fibo numbered extensions.

http://www.accrete.com/fx-posts/tgle...ACDexample.gif

Cheers,
Thom
  • #7
  • Oct 12, 2006 12:55am Oct 12, 2006 12:55am
Quoting Captain Piptastic
I will agree with you that many new traders (hell, maybe old traders too) pull targets and stop loss numbers right out of their ass. However, fixed S/L and Targets can be tested and statistics can be generated for them just like any other exit methodology.

You may indeed find that using the same indicator(s) and an opposing signal is the best exit. There are a lot of ways to trade, however, and my experience has been that some methods can do very well (even better) with fixed targets as compared to indicator exits.

I will say that fixed targets tend to do better with a breakeven floor amount and some type of trailing stop.
An excellent answer Captain Piptastic,

A lot of traders who would never try to prematurely pick the tops (and bottoms) to enter a trade, often try to pick the tops (and bottoms) to exit a trade. The result is to then see the market keep rising (or falling) with the subsequent loss of pips.

It is far easier to use a trailing stop and as the profits get bigger and bigger (that is, the elastic being pulled is getting more and more stretched) then make the trailing stop smaller and smaller. It sure works for me.

Just my 2 cents worth.
  • #8
  • Oct 12, 2006 1:12am Oct 12, 2006 1:12am
Quoting accrete
My 2 pips on the topic of exit picking...

In Tom DeMark's book "The New Science of Technical Analysis" he talks about some of his favorite moving averages. For my example/thoughts i'll take two of his favorites, the 13 and 55. Now he dosn't use them in the typical "Cross" to enter a trade, rather he takes notice the first day both change directions in unison. Anyway, we are talking exit picking here, not entries, so i ramble. For a simple excersize (though this will take some quality manual backtesting/study time), take these same two MAs and take note of their location in time and space where they cross. Now instead of using Vega's moving Median MA with Fibo-Numbered-Extensions, take those same FNExt from the point of the cross and see if there is a correlation between the numbers and stalls/consolidations/reversals. Maybe you begin to see some patterns taking place on a 30m or 60m TF with these MAs? Do some more research, see some more correlations?? In the long run there are patterns that will be recognized by using the above method and one can base exit strategies/rules by these studies that will result in greater gains than simply picking past S/R or XX number of profit-pips.

This is how i choose my exits, and it is profitable rather than the moving fibo TP areas off a median MA.

Here is a quick post script insert of an EA based trade i've been experimenting with that did what i discussed above. You can see that the TP levels were set based off the MA intersection, then at that moment in time, the levels were calculated and TPs set in stone more or less, and the trade was then let run it's course taking profit at the Fibo numbered extensions. This is in contrast to simply visualizing where the currency is in reference to the moving median fibo numbered extensions.

http://www.accrete.com/fx-posts/tgle...ACDexample.gif

Cheers,
Thom
can you post your EA
{Link Removed}]
  • #9
  • Oct 12, 2006 7:01am Oct 12, 2006 7:01am
Quoting Captain Piptastic
You are assuming an awful lot here. You know a move is coming? You know there is a certain probability of a move based on your backtesting. Why can't you know that a certain percentage of those moves go x number of pips?
This is very true. I have tested various systems that use fixed stops and targets and over time, nothing else could beat them. In a data series there is an average target and loss. If you adhere to those values they will grow your equity.

The other nice thing about fixed values is that you know your risk:reward going into the position and you are usually out of the position much sooner than someone who has to monitor the market for their exit conditions.
  • #10
  • Oct 12, 2006 8:28am Oct 12, 2006 8:28am
I really don't get the "peel the bananna" type exit strategy...

where somone enters a position with, say, 3 lots and exits at 3 different targets. e.g. lets say they got out at 30, 60 and 90 pips... so they claim they made 180 pips! but really they made 60 pips / lot. I mean by this logic I could say I'm a millionaire (but I guess Turkish Lira doesn't count )

no matter how many angles I look at this from, for me it doesn't add up. Sounds better to me just to take a good trade and go "all-in" until you decide to exit. That way you only need a 60pip move to make 60 pips, not a 90 pip move.

and I don't see how it can make a trade any less risky.
  • #11
  • Edited Oct 13, 2006 2:28am Oct 12, 2006 8:49am | Edited Oct 13, 2006 2:28am
Quoting ycomp
I really don't get the "peel the bananna" type exit strategy...

where somone enters a position with, say, 3 lots and exits at 3 different targets. e.g. lets say they got out at 30, 60 and 90 pips... so they claim they made 180 pips! but really they made 60 pips / lot. I mean by this logic I could say I'm a millionaire (but I guess Turkish Lira doesn't count )

no matter how many angles I look at this from, for me it doesn't add up. Sounds better to me just to take a good trade and go "all-in" until you decide to exit. That way you only need a 60pip move to make 60 pips, not a 90 pip move.

and I don't see how it can make a trade any less risky.
I'm really glad you brought this up. That is a huge pet peeve of mine. I'm convinced that most of the people that do this, do it so they can brag about their awesome pip count. Someday I'll start a thread titled "Pip counts don't mean $%&*!" and rant about it.

Regarding scaling out:
I admit that I don't like this idea very much. It almost always lowers the theoretical maxium you can gain. However, it can smooth out an equity curve and have psychological benefits.

Mark Douglas talks about this benefit in "Trading in the Zone". The first time I read it I rejected it outright. Sometime later when I was struggling with one of my methods I revisited it. Taking partial profit did really help me "get over the hump" and let my trades run further.

These TP's were predetermined points that I knew had a high probablity of being reached. I knew those probablities because I backtested them just like I backtested other exit strategies. They were not random numbers. From there I could employ various TA techniques for the rest of the position.

I'm not trying to get off topic here. Andrew wanted to discuss the "best" exit methods. If we consider best we must consider all the benefits and disadvantages of various exit methods not simply the theoretical ending balance.

BTW, I should have said this earlier, thanks to andersenws for starting this thread. Exits are not discussed enough. They seperate the consistent winners from the consistent losers.
Secret Identity: Bill Sullivan
  • #12
  • Last Post: Oct 12, 2006 8:54pm Oct 12, 2006 8:54pm
Thanks for the replies guys. I definitely learned some more about exits, which I agree is not talked about enough. I did have a specific system/EA in mind when I started this thread, and that was what got me thinking about this topic in the first place. I am now in the middle of a forward test for this EA and will probably start a thread later with some tests of it.

Thanks again,
andersenws
"Youth is the trustee of prosperity." - Benjamin Disraeli
0 traders viewing now