Accuracy? Who Needs It...

[jonbig04 10]

Since I'm always droning on about this I figured I would just make a thread dedicated to my droning.

 

Cut your losses and let your winners run. Cliche? No doubt, but that doesn't make it any less important.We all know that there are about a billion ways to trade profitably. So really, we can't say that one method is better than the other. Unfortunately, that doesn't help newbies (like myself) very much. This site is full of information. How to trade, what to trade, who to use to trade etc. But it sees like the vast majority of it is dedicated to accuracy. Making the right call, the solid play, the perfect trade. Who cares? Accuracy can only be discussed in the context of risk:reward. If your risk is 1 point and your reward is 125 a person with 1% accuracy can make all kinds of money. Conversely a person with a risk of 90 and a reward of 1 can, with 99% accuracy, make all kinds of money. With that in mind why are so many more concerned about their accuracy than their reward. There are equal parts of the profitable traders equation.

 

So how do we let our winners run? Well thats what I hope for this thread to be about. Many people, when in a winning trade, are more concerned about protecting their profit than letting the thing actually run. Why? Because they are scared it will come back down to bite them. I maintain that most scaling out is due to FEAR, not expectancy. Of all the profitable traders I know, this is the biggest difference I see between their trading and the trading of us noobs. We can't be afraid to let our winners run.

 

One thing I am doing is looking at a larger time frame chart. I no longer watch the 1 min or 1k vol (nq) chart. I use a bigger chart. If the market is fractal, why wouldn't I? I don't want to hold overnight for personal reasons, so I have the biggest chart that will still show me intraday trades. Why catch smaller moves intraday if I can catch bigger ones? If I'm going to let my winner really run I need to see a bigger picture. One thing I will say is, though I increased my time frames, I did NOT increase the size of my stop. it's still intraday. If I was swing trading then the stop would have to be increased, but for the size of stops I see everyday in chat, we can be looking at much bigger time frames without increasing our risk.

 

 

Anyways, I hope some other people will chime in on the problem of letting our winners run. True, scaling out at pivotal areas can be a good idea, if something changes...yea I get it. But thats not my concern, because I know damn well that the majority of scales intraday are because we are scared something MIGHT happen, not because of something we are seeing right then in PA.

[diablo272 10]

So you're basically saying that positive expectancy is important?

[jonbig04 10]

lol, well naturally. What I'm saying is that we focus much more on the accuracy side of the expectancy equation than we do the reward size side.

[zdo 20]

jonbig04

 

‘Letting winners run’ can not be ubiquitously applied to all systems!

 

It MUST be applied to trend following systems because capturing outliers IS the edge

of trend following systems. So, in trend following systems, scaling out is not worthwhile.

 

But for ALL other types of systems only varying gradients of ‘letting winners run’ is appropriate. And scaling out, quickly or more gradually per the characteristics and statistics of the non trend following system, does have efficacy.

 

Those that don’t get this might as well be straw sucking ‘daylight DMT’ every morning…

even though having this knowledge does hardly anything to help conquer those fears / biases you discussed in your post.

[Head2k 10]

It's not black and white. Accuracy is important, of course. The more accurate you are the less risk you undertake and/or the more winners you have compared to number of losses. And if you can increase your accuracy while not decreasing your avg. win / avg. loss, then it definitely helps you.

And concerning time frames, a larger bar interval chart is fine for watching what's the situation on a larger time frame, but you can use even 1 tick chart and still trade a large time frame. Time frames or wave sizes - or whatever you call it - are interrelated, and if you understand these relations you can use a fast chart and accurate (high prob. and little risk) entry into a large time frame trade.

As for scaling out, I see the main purpose in protecting achieved gains for the case that something unexpected happens, and in letting winners run longer if price goes through the estimated target. If you enter a large time frame trade, or in other words if you want to ride a big wave, then this wave likely won't be smooth and what is more important, you never know whether you really entered on THE wave, whether your target estimation is correct, whether S/R will be found where you expect it and whether it will hold or not.

[jonbig04 10]

jonbig04

 

‘Letting winners run’ can not be ubiquitously applied to all systems!

 

It MUST be applied to trend following systems because capturing outliers IS the edge

of trend following systems. So, in trend following systems, scaling out is not worthwhile.

 

But for ALL other types of systems only varying gradients of ‘letting winners run’ is appropriate. And scaling out, quickly or more gradually per the characteristics and statistics of the non trend following system, does have efficacy.

 

Those that don’t get this might as well be straw sucking ‘daylight DMT’ every morning…

even though having this knowledge does hardly anything to help conquer those fears / biases you discussed in your post.

 

"Trend following" is tough to define. After all, which trend do you mean? For me personally I consider myself a trend follower-of the last few days or weeks. During the day I'm sometimes counter-trend, and on entry I ALWAYS am counter trend. So am I a trend follower or not?

 

For someone just trading reversals, lets say on a week long chart, I think holding your winners is even more important. Because (IMO) those are tough and when you do nail one, there is lots of profit potential.

[jonbig04 10]

It's not black and white. Accuracy is important, of course. The more accurate you are the less risk you undertake and/or the more winners you have compared to number of losses. And if you can increase your accuracy while not decreasing your avg. win / avg. loss, then it definitely helps you.

 

 

 

true, accuracy is of course necessary-my point is simply that its only HALF of the equation.

 

 

 

And concerning time frames, a larger bar interval chart is fine for watching what's the situation on a larger time frame, but you can use even 1 tick chart and still trade a large time frame. Time frames or wave sizes - or whatever you call it - are interrelated, and if you understand these relations you can use a fast chart and accurate (high prob. and little risk) entry into a large time frame trade.

 

 

Right, this is what I think the vast majority of us do to some extent. (identify trend, zoom in for entry).

 

 

As for scaling out, I see the main purpose in protecting achieved gains for the case that something unexpected happens, and in letting winners run longer if price goes through the estimated target. If you enter a large time frame trade, or in other words if you want to ride a big wave, then this wave likely won't be smooth and what is more important, you never know whether you really entered on THE wave, whether your target estimation is correct, whether S/R will be found where you expect it and whether it will hold or not.

 

True, no doubt sometimes the unexpected does happen. the other day I saw my +8ES turn into a -1. If you start letting your winners really run, this WILL happen (most move their stop to BE, I dont). So that does happen, but usually, if you are that far ahead, the move won't reverse in the middle of nowhere. So what we have is all these people who are so scared of being subject to the psychological anguish of watching a +10ES ( or more, etc) turn into a BE that they scale out or exit entirely, when the majority of the time the win was there-we are just to afraid of what it will feel like if we see a big gain disappear so we cut our winners short.

 

When that happened to me the other day someone remarked "I would never let a gain like that turn into a loss or BE". Thats probably true, but from what have seen most people would have never held long enough to see the +8 to begin with.

 

What I am getting at here is this: I think that a lot of people are leaving a lot of money on the table by cutting their winners short. To make up for this people have to increase accuracy, and they are struggling to find a way to do that. When the only thing holding them back from really holding their winners (thus increasing their expectancy) is fear. Fear of watching that gain disappear, because once they have a winner they are so scared of losing it, that emotion takes control.

[mezameo 10]

I think this is an excellent way of putting it, letting winners run is a question of risk tolerance. Perhaps even more important than our equity balance is our patience threshold.

 

Without patience to let a trade signal, or the market-- if you are a pure price action type of trader--take you out at a profit or a loss, your risk of being wrong will never even come in to play. And certainly when you look at historical charts of all the great trades in our lifetime, we can be sure that without question, all great trades have tested our patience.

 

Look at a monthly chart of the EURO for example. Right now the Monthly chart is in a massive phase of consolidation. The trend is clearly down however. But who among us has the patience to hold on to a winner for more than a few days or weeks to see it really pay off?

[jonbig04 10]

I think this is an excellent way of putting it, letting winners run is a question of risk tolerance. Perhaps even more important than our equity balance is our patience threshold.

 

Without patience to let a trade signal, or the market-- if you are a pure price action type of trader--take you out at a profit or a loss, your risk of being wrong will never even come in to play. And certainly when you look at historical charts of all the great trades in our lifetime, we can be sure that without question, all great trades have tested our patience.

 

Look at a monthly chart of the EURO for example. Right now the Monthly chart is in a massive phase of consolidation. The trend is clearly down however. But who among us has the patience to hold on to a winner for more than a few days or weeks to see it really pay off?

 

 

Good points. I'm not even advocating swing trading, I'm just advocating larger time frames and..well...fearless holding. You are absolutely right, the best trades are the ones that last all day long testing and trying us through every minute. People scale to avoid this, I think.

 

Another thing I am curious about is something I read here on TL. Its that we scale because the odds have changed, out of our favor. How does that work? If I enter on say a potential double-bottom, and price starts taking off, isn't that putting the odds more and more in my favor?

 

When we reach a profit, should our reaction be more along the lines of holding because it looks like the odds are more on our side as opposed to being scared that something might change and bailing out?

 

Another thing. I read in chat that some people exit when buyers ( if they are short) or sellers (if they are long) might step in. Makes sense, but let me submit this. If you enter a trade on NQ and you scale out at +5 because you think the other side "might step in", you should never have taken the trade in the first place. (Keep in mind that I'm referring to the way the majority of us trade. I'm sure there are scalpers who make bank by netting +5s all day, but it doesn't help anyone if we constantly advocate every kind of style). If there is a place where the other side might step in, i sure as hell am not going to enter 5NQ below it (or above it). Thats shitty R/R. We need to enter in places where our trades have room to move. If there isn't room, than maybe we shouldn't take the trade.

[atto 20]

From someone with a less than 50% accuracy, I mostly agree with what you're saying. However, I do disagree about scale outs. If you know much about Blackjack card counting, you know that your win rate is still under 50%; however, you simply increase your bet when you have the statistical advantage.

 

This is where concepts such as the Kelly criterion come from. This applies directly to trading: when your edge is greatest, trade the biggest size. Most simply, this is your entry: elsewhere, your edge is 0 or slightly negative. When you enter (with a statistical edge, of course), you've calculated that you have a positive edge, so you put on size.

 

The true idea behind scaling out is to manage your position with respect to your current edge. This assumes an important thing: you can roughly calculate your edge. For myself, I don't have the exact numbers; some do (Hlm probably has a very good idea of his actual edge). However, let's enter a hypothetical long that goes well. Eventually, we approach established resistance. Couldn't you argue that holding a long into resistance has a lower edge than, say, holding a long that just bounced off support?

 

Or, what if, during your long, price fails to make a higher high, and makes a lower low? That's the first sign of a trend reversal. Your edge is less.

[Hlm 10]

Another thing. I read in chat that some people exit when buyers ( if they are short) or sellers (if they are long) might step in. Makes sense, but let me submit this. If you enter a trade on NQ and you scale out at +5 because you think the other side "might step in", you should never have taken the trade in the first place. (Keep in mind that I'm referring to the way the majority of us trade. I'm sure there are scalpers who make bank by netting +5s all day, but it doesn't help anyone if we constantly advocate every kind of style). If there is a place where the other side might step in, i sure as hell am not going to enter 5NQ below it (or above it). Thats shitty R/R. We need to enter in places where our trades have room to move. If there isn't room, than maybe we shouldn't take the trade.

I will reply since I am pretty sure you are talking about me.

Also atto already made some good points.

 

A few quick points.

 

• Market is made of up different sized trends/swings/etc (fractal "like")
  
• To create a high r:r one looking for least resistance through two or more of these.
  
• Potential S/R is the area where trends are tested.
  

It's a probabilities game...you don't "know" when the other side will step in. What you can estimate is the higher/lower areas along the way. You talk about one giving room for their trade to move, but don't you use very small stops and will take multiple ones until you hit that large trade (please correct me if I am wrong)? There are always smaller areas above and below your entries where the other side may step in (especially if you are entering on pullbacks). If someone has a very high r:r this is why they many times have to take a few stops before hitting a full target. Maybe an example of what you mean on a chart would be helpful. Odds move in similar waves as price. As price approaches an area of S/R (say an opposite direction trend that needs to fail in order for your trend to continue) the odds decrease that price will continue through. The further price gets on the other side of that area the probabilities increase that it will continue to the next area of S/R. One also has to remember that it's not necessarily a line in the sand. As you move into larger trends, the volatility and deviation increases.

 

For an example let's take the concepts of Value Areas or MP. In very simple terms, there is an 80% chance that price will rotate to the other side of a value area if there is lack of interest found outside the starting side. One can keep taking shorts at rejection outside the value area and potentially keep getting stopped out as price finds little interest back inside and tests outside again. Eventually price may fall to the other side and profit is made though it's not guaranteed by any means...it's just a probability. Another way it can be traded is to take the short at rejection, but take some off on the test back inside with rest at b/e. One will end in profit either way and multiple attempts can be made (profit made during the 20% time). One could also move their stop to net b/e (room for more tests outside w/out being stopped out) and walk away. None are necessarily wrong or right (though there will be differences), what's key is knowing the areas the market is speaking.

 

In my opinion, it's not so much about scaling out or not. It's about understanding the different sized trends you area dealing with and building your money management around them. It's about keeping everything dynamic and letting the market tell you what to do.

 

A few different MM examples...

 

• Trail stop (no scale out) based on these areas.
  
• Scale out a large portion at first area of S/R and move stop to net b/e (giving much more room).
  
• Scale out at every area along the way.
  
• A mixture of the above.
  

In the long run, many times the bottom line will be very similar. What's different is the draw down and shape of the PnL curve that creates that bottom line.

 

Good thread we have going here, hopefully it can stay clean and on track.

[Hlm 10]

As price approaches an area of S/R (say an opposite direction trend that needs to fail in order for your trend to continue) the odds decrease that price will continue through.

After reading what I wrote, I realized that this sentence might not have been clear. The odds don't necessarily decrease at the area of S/R...it decrease at the current price as you get closer to the potential S/R. In other words there is a set gradient of probabilities that price moves along. I used the term "necessarily" above because depending on the price action in front of the S/R, there can be changes in the summation of probability at the potential S/R. Hopefully that is a little more clear.

[jonbig04 10]

From someone with a less than 50% accuracy, I mostly agree with what you're saying. However, I do disagree about scale outs. If you know much about Blackjack card counting, you know that your win rate is still under 50%; however, you simply increase your bet when you have the statistical advantage. This is where concepts such as the Kelly criterion come from. This applies directly to trading: when your edge is greatest, trade the biggest size. Most simply, this is your entry: elsewhere, your edge is 0 or slightly negative. When you enter (with a statistical edge, of course), you've calculated that you have a positive edge, so you put on size.

 

 

You may have gone over my head here lol. I don't understand how, when PA is doing what you expect, your advantage decreases. Lets say you are playing a standard head and shoulders pattern. You enter at what you anticipate to by the forming of the next shoulder, your trade is now in the green by say 5NQ. Then the neckline breaks. Wouldn't your edge increase? If you are looking for X (in price action) and you begin to see it, aren't you in an improving position?

 

The true idea behind scaling out is to manage your position with respect to your current edge. This assumes an important thing: you can roughly calculate your edge. For myself, I don't have the exact numbers; some do (Hlm probably has a very good idea of his actual edge). However, let's enter a hypothetical long that goes well. Eventually, we approach established resistance. Couldn't you argue that holding a long into resistance has a lower edge than, say, holding a long that just bounced off support?

 

IMO all R and S weren't created equal. I get my levels from a large time frame chart. lets say I enter a long based of a support level from that chart, it it runs into intraday R. I am betting that since my S is from a larger time frame, it is more significant and hopefully stronger than anything the intraday can throw at me. Now if by established R you mean R that was ALSO found on a large time frame chart, than absolutely scale or exit. But for me those levels are roughly 20-30NQ apart-so holding until then is the idea. I try not to let intraday p/a scare me.

 

Or, what if, during your long, price fails to make a higher high, and makes a lower low? That's the first sign of a trend reversal. Your edge is less.

 

I think this depends on which chart you're looking at. For me personally, I dont trade between S/R (large time frame). Why? Because I'm not good enough. There are too many shakeouts and fakes. I trade the extremes and whatever happens in the middle is complete bs that I TRY not to pay attention to.

[jonbig04 10]

I will reply since I am pretty sure you are talking about me.

Also atto already made some good points.

 

A few quick points.

 

• Market is made of up different sized trends/swings/etc (fractal "like")
  
• To create a high r:r one looking for least resistance through two or more of these.
  
• Potential S/R is the area where trends are tested.
  

It's a probabilities game...you don't "know" when the other side will step in. What you can estimate is the higher/lower areas along the way. You talk about one giving room for their trade to move, but don't you use very small stops and will take multiple ones until you hit that large trade (please correct me if I am wrong)? There are always smaller areas above and below your entries where the other side may step in (especially if you are entering on pullbacks). If someone has a very high r:r this is why they many times have to take a few stops before hitting a full target. Maybe an example of what you mean on a chart would be helpful. Odds move in similar waves as price. As price approaches an area of S/R (say an opposite direction trend that needs to fail in order for your trend to continue) the odds decrease that price will continue through. The further price gets on the other side of that area the probabilities increase that it will continue to the next area of S/R. One also has to remember that it's not necessarily a line in the sand. As you move into larger trends, the volatility and deviation increases.

 

For an example let's take the concepts of Value Areas or MP. In very simple terms, there is an 80% chance that price will rotate to the other side of a value area if there is lack of interest found outside the starting side. One can keep taking shorts at rejection outside the value area and potentially keep getting stopped out as price finds little interest back inside and tests outside again. Eventually price may fall to the other side and profit is made though it's not guaranteed by any means...it's just a probability. Another way it can be traded is to take the short at rejection, but take some off on the test back inside with rest at b/e. One will end in profit either way and multiple attempts can be made (profit made during the 20% time). One could also move their stop to net b/e (room for more tests outside w/out being stopped out) and walk away. None are necessarily wrong or right (though there will be differences), what's key is knowing the areas the market is speaking.

 

In my opinion, it's not so much about scaling out or not. It's about understanding the different sized trends you area dealing with and building your money management around them. It's about keeping everything dynamic and letting the market tell you what to do.

 

A few different MM examples...

 

• Trail stop (no scale out) based on these areas.
  
• Scale out a large portion at first area of S/R and move stop to net b/e (giving much more room).
  
• Scale out at every area along the way.
  
• A mixture of the above.
  

In the long run, many times the bottom line will be very similar. What's different is the draw down and shape of the PnL curve that creates that bottom line.

 

Good thread we have going here, hopefully it can stay clean and on track.

 

Great post. I personally don't take multiple shots at entries. I try to nail it on the 5 sec and if I don't, I pay the price. But I see your point. I think it comes down to what you said near the end and what I mentioned in the above posts. Its about time frames. Instead of watching the shaky intraday I try to locate the larger trend and catch a small part of that. About the lines in the sand, I somewhat disagree. To me NQ seems to bend and move s/r much more than ES. I don't know why, but there are hardly any lines in the sand on NQ. ES though has lines in the sand to some degree. I'm actually doing a study on it right now and will post my results, but many times ES respects a level to the point, eg 836. To me 836 on ES is better than sex. yea, I said it lol. The other day you took a long from 836. I missed the boat on the 5 sec, but was glad to see you caught it. With 836 such an important level to me there is a decent chance that it will win the fight with intraday R, where you scaled out. So why not push it? With worst case a BE and best case a +15ES and the knowledge that it happens fairly often, I grab my balls and hold it. A typical bounce off of that important of a level ( i dont know it off the top of my head), is quite a bit more than 4ES or 5ES. So why not just exit when the larger time frame makes a new HH or LL, or you reach what you think to be an equally important level. If the market moves quickly from S to R then the small psychological damage you receive from watching a +3ES turn into a b/e will be nothing compared to the couple times you book the +10 or +15ES. Thats just my opinion though.

 

You guys know way more about this stuff than me of course. From reading the chat though I notice that some people aren't yet profitable in SIM, so I'm just offering my $0.02 to them as well as anyone else.

[Hlm 10]

This is another thread that one must not assume that everyone is on the same page.

 

Much of it depends on...

• How you define your S/R
  
• Expectations at S/R
  
• How you define your "time frames"
  
• How many "time frames" go into your analysis
  

To me NQ seems to bend and move s/r much more than ES. I don't know why, but there are hardly any lines in the sand on NQ.

I don't really see where you are coming from with this comment. Maybe you are trying to force the same sized trends/tfs seen on the ES onto the NQ in a cookie cutter way? You mention that you are doing a study on it right now, I am very interested to see what you come up with. Please let us know.

 

...so I'm just offering my $0.02 to them as well as anyone else.

As you should...it's always good to see other peoples views. Not only to help others, but also to help ones self. I have learned so much from having to analyze and answer questions asked about certain aspects of my trading.

 

The one point where everyone should be on the same page is that if one chooses to scale out, it should not be based on emotion. Though it may work for the short-term, it will more than likely hurt you in the long-term.

[jonbig04 10]

This is another thread that one must not assume that everyone is on the same page.

 

Much of it depends on...

• How you define your S/R
  
• Expectations at S/R
  
• How you define your "time frames"
  
• How many "time frames" go into your analysis
  

I don't really see where you are coming from with this comment. Maybe you are trying to force the same sized trends/tfs seen on the ES onto the NQ in a cookie cutter way? You mention that you are doing a study on it right now, I am very interested to see what you come up with. Please let us know.

 

As you should...it's always good to see other peoples views. Not only to help others, but also to help ones self. I have learned so much from having to analyze and answer questions asked about certain aspects of my trading.

 

The one point where everyone should be on the same page is that if one chooses to scale out, it should not be based on emotion. Though it may work for the short-term, it will more than likely hurt you in the long-term.

 

 

What I meant about NQ is that it seems to bend s/r much more than ES. in part because of the small tick size. If I pulled up morning reversals at globex, NQ would likely reverse sometimes at ghigh (or glow), but usually it will be 2-6 points above it, where as ES will stop dead in its tracks more often. If I see level I want on ES, I can place a limit there and still have a decent chance, its tough to do that on NQ.

 

Your last point is really what its all about. I'm just advocating looking at larger trends, as they tend to chop MUCH less than trying to make sense out of the intraday charts (1 min etc) and basing your targets accordingly. Enter at important levels on the larger chart so that your stop is protected (ideally) AND you can target the next level on the chart which will be a good number of points away (you dont have to worry about buyers or sellers stepping, because you are trading something bigger). IMO you can do this with the same size stops that i see people using everyday in chat when they are trading intraday charts. So why not increase your reward? The only hard part is actually holding through the sometimes scary intraday action, and no I don't mean necessarily swing trading (I close all my trades at EOD).