Expectancy is always zero | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

What is Expectancy? 13 replies

System Expectancy - Monte Carlo Simulation 5 replies

trade expectancy help 0 replies

Expectancy question for experienced traders 2 replies

Random Market vs Set Expectancy 8 replies

Attachments: Expectancy is always zero
Exit Attachments

Expectancy is always zero

  • #1
  • First Post: Dec 13, 2007 8:57pm Dec 13, 2007 8:57pm
Actually, without some definite edge, expectancy is always negative because of spreads and commissions. Here's why.

Most traders will be familiar with the standard expectancy equation:

W = win rate
G = average Gain (or predefined TP if there is one)
L = average Loss (or predefined SL point if there is one)

Expectancy = WG - (1-W)L

This is like trading math 101 but it is very compelling to new traders who think that it is the key to sure consistent success. "After all," they think, "if I just cut my losses to 10 pips, and let my trades run to 30 pips, then even if I only win 30% of the time then I would still be ahead!"

That is of course because (0.30 x 30) - (0.70 x 10) = 9 - 7 = 2 pips/trade.

Wow! Any dolt can be right only 30% of the time, so all I have to do is make a million trades this year and I'll be up 2 million pips! Right? Right????

Ummm, no. :

Let's get beyond trading math 101 to the next step. What's wrong with the train of thought above is that the win rate is actually a function of your risk/reward ratio.

Suppose you set your stop at 1 pip and your TP at 800 pips on every trade. How many do you think you'd win? Your win rate would be virtually zero. Now if you reversed the situation and set your TP at 1 pip and your stop at 800 pips then your win rate would be very close to 100% If your SL and TP are at the same distance from your entry, then your win rate should be about 50%. In fact, your win rate (all else being equal) should be equal to the distance to the stop divided by the total distance from the stop to the target.

For example if your SL is 20 pips from entry and your TP is 80 pips from entry, then your win rate on trades like that should be around 20%. If you set a stop at 20 pips and a TP at 5 pips then your win rate should be about
20 / (20 + 5) = 20/25 = 80%

So I'm saying that in terms of the variables in the standard "expectancy" equation, the win rate is actually equal to: L / (L+G)

So let's plug that into the expectancy equation and see what we get shall we? We started with:

E = WG - (1-W)L

but now we know that W = L / (L+G) so...

E = LG / (L+G) - L + L^2 / (L+G) note: L^ is "L squared."

multiplying the middle term on the right by (L+G)/(L+G) which is just 1:

E = LG / (L+G) - L(L+G) / (L+G) + L^2 / (L+G)

combining into one fraction with a common denominator:

E = (LG - L(L+G) + L^2) / (L+G)

E = (LG - L^2 - LG + L^2) / (L+G)

E= 0 / (L+G)

E = 0


And that in a nutshell is why trading can be difficult. Just depending on old pat bromides like "cut your losses short and let your profits run," or "you can't go broke taking profits," although they sound good will not really get you anywhere.

Sure, if you cut your losses short, you know what you get? A lot of little losses which just about equal your very infrequent big wins. If you follow the opposite advice that "you can't go broke taking profits," and decide to use wide stops and tight targets guess what? You get a lot of little profits that just about make up for that huge loss you take every so often.

All of these cute slogans do nothing more than represent different points along the risk/reward spectrum, but traders should be aware that each of those points has an expectancy of about zero. And that's not including costs. So forget the cute slogans. The key to trading successfully is to find an edge that moves that E value slightly to the positive side of your costs.
  • #2
  • Dec 13, 2007 9:14pm Dec 13, 2007 9:14pm
Quoting capitalist88
Actually, without some definite edge, expectancy is always negative because of spreads and commissions.
............
The key to trading successfully is to find an edge that moves that E value slightly to the positive side of your costs.
if that's the case, Swap will solve the problem....
  • #3
  • Dec 13, 2007 9:32pm Dec 13, 2007 9:32pm
Morning,

I love numbers aswell, so I found your post interesting. Something some new traders fail to take in to consideration is the 'true' impact of spread, its very simple but very limiting, in escence spread is actally double(or is it)......

A trader has a 3pips spread on a pair, if price was to move 5pips in ones favour then you have 2pips floating, whereas... if price moves 5pips against the trader then you have -8pips floating.
So in that example, thesame pip move gives 4times the loss compaired to gain.
This simple fact makes one think about the serious pitfalls of 'scalping' for small profits fom the markets.

But im sure you already knew that, posted for others to take note though.

Leon
Living the adventure in my head.
  • #4
  • Dec 13, 2007 10:07pm Dec 13, 2007 10:07pm
Yes, that may be true to a certain extent, if the TP is further from entry than the SL, ceteris paribus, the SL will be hit more often than the TP in a market with equal probability of moving either way. The greater the R:R ratio the less likely that your TP will be hit and the more likely that the SL will be hit. That is why we identify S+R levels, potential zones of reversal, trends etc., as you say; to increase the value of E.

Still, I dispute some of the assumptions in your model - although everything balances out nicely on paper, they simply do not hold in reality. Random entry systems (toss a coin, enter) have been proven to be successful in the long run when implementing the 'cut your losses, let your profits run' adage which disproves your argument. If you were to map the probabilities I'm fairly certain it would not be a normal distribution, once price moves in a certain direction, it often keeps moving in that direction in a trending market. When you're right, you've got a good chance of TP being hit if you have been scientific and realistic about your TP placement (which is a can of worms in itself), and when you're wrong, you're wrong and you find out soon enough. Greater minds than mine will give you a better explanation as to why this is but the evidence speaks for itself.

Take a look at Dr K Van Tharp's random entry work. I think there's a thread here where somebody has experimented with the toss a coin technique. I'll see if I can dig it up.
  • #5
  • Dec 13, 2007 10:36pm Dec 13, 2007 10:36pm
Hello Capitalist,

First let me state this, you confused me when you got into the formulas. So if I am saying something you eluded to and I didn't catch it, please state so.

Quoting capitalist88

If your SL and TP are at the same distance from your entry, then your win rate should be about 50%.

.
I greatly enjoyed your post and certainly look forward to more in the future. I was wondering if you may be able to shed light, and your opinion, on a trending environment?

**My Opinion**

Your above comment would appear true, though I have found that, at least in my experience, that it isn't. Even this scenario, as simple as it sounds, would require fairly unusual and sought after circumstances. Allow me to try and explain my position. One would have to be trading in the middle of a range for this to be in effect. A pair would need to be stuck in that range, have absolutely no measurable pattern of how it behaves in that range and have no external or changed momentum characteristics. Then, if a trade were taken in the middle, an equal TP and SL would more than likely bring the % to within +/- 1. One would assume you could find these conditions if searching hard enough.

**Question**

Having stated that, how would you examine the statistics of the previously mentioned trending market? For instance, the CAD/JPY in 335 days has gone up 9%. This is just a measly .0267% a day (assuming 9% = 9.0 not .09 as one would use in other equations). However, this can be factored to a specific market condition and be applied to increase % could it not? Or would this be subject to randomness somehow? Would placing a 50 TP and 50 SL still be of very little concern during an obvious trend downwards or upwards? Perhaps the market could change at any moment and swing in any direction. Afterall, hindsight is always 20/20. This is something I am currently exploring, though I admit I am having a tough time finding a formula to use. I would very much like to be able to factor the win % of a set TP with an equal SL in a given market condition. Like the above CAD/JPY. I am hardly a math wizard, in fact, the only thing limiting me from graduating college with honors is a college algebra course. Go figure.
The market is my nation. Traders, my family. Hello, brothers and sisters!
  • #6
  • Dec 13, 2007 11:33pm Dec 13, 2007 11:33pm
Jay is correct: the 50% hit rate given 1:1 risk/reward is not true.

It would only be true if the market was a random walk in the space of possible prices but the market is not a random walk model. This has been mathematically proved since the 90's in various rigorous studies in the Journal of Finance..etc. The market is only rarely random, especially when there is no speculative activity. Your probability of success is a function of your risk/reward, yes, but there are also other unknown factors. Just because they are unknown does not mean they are not there. Using previous information (e.g patterns, order flow information) you can have at least some of those factors on your side.

Hope this helps.
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #7
  • Dec 14, 2007 12:05am Dec 14, 2007 12:05am
Quoting Gwan
if that's the case, Swap will solve the problem....
I'm a big fan of the carry trade so I'm with you on this!

  • #8
  • Dec 14, 2007 12:10am Dec 14, 2007 12:10am
Quoting Leonlorenzo
Morning,

I love numbers aswell, so I found your post interesting. Something some new traders fail to take in to consideration is the 'true' impact of spread, its very simple but very limiting, in escence spread is actally double(or is it)......

A trader has a 3pips spread on a pair, if price was to move 5pips in ones favour then you have 2pips floating, whereas... if price moves 5pips against the trader then you have -8pips floating.
So in that example, thesame pip move gives 4times the loss compaired to gain.
This simple fact makes one think about the serious pitfalls of 'scalping' for small profits fom the markets.

But im sure you already knew that, posted for others to take note though.

Leon

Another excellent point that should be on the "list" for newer traders. As you probably know I've been I've been experimenting with ultra short TFs myself a lot lately and the limiting factor really is the spread. I've been reading through the entire 265 page LinuxTroll Simple Scalping thread to see how people did with it. Interesting stuff.
  • #9
  • Dec 14, 2007 12:33am Dec 14, 2007 12:33am
Quoting birdt
Yes, that may be true to a certain extent, if the TP is further from entry than the SL, ceteris paribus, the SL will be hit more often than the TP in a market with equal probability of moving either way. The greater the R:R ratio the less likely that your TP will be hit and the more likely that the SL will be hit. That is why we identify S+R levels, potential zones of reversal, trends etc., as you say; to increase the value of E.

Still, I dispute some of the assumptions in your model - although everything balances out nicely on paper, they simply do not hold in reality. Random entry systems (toss a coin, enter) have been proven to be successful in the long run when implementing the 'cut your losses, let your profits run' adage which disproves your argument. If you were to map the probabilities I'm fairly certain it would not be a normal distribution, once price moves in a certain direction, it often keeps moving in that direction in a trending market. When you're right, you've got a good chance of TP being hit if you have been scientific and realistic about your TP placement (which is a can of worms in itself), and when you're wrong, you're wrong and you find out soon enough. Greater minds than mine will give you a better explanation as to why this is but the evidence speaks for itself.

Take a look at Dr K Van Tharp's random entry work. I think there's a thread here where somebody has experimented with the toss a coin technique. I'll see if I can dig it up.
Hi birdt,

I'm familiar with Van Tharp's work and the random entry experiment. Your point is a valid one and has to do with such issues as whether price movement distributions have fatter tails ("black swan" events in common usage), infinite variance, etc. Basically if gain is potentially unlimited while risk is limited then there's an expectancy advantage there.

You and the two posters after you make some very good points with which I agree, and I commend you all on your depth of knowledge on these subjects. Let me clarify though that my intended audience in creating the thread was newer traders who may still be at the point where they think all they need to do is set up a 1:3 RRR and life will be golden.

So while I agree with you and the two posters after you in many respects, you guys are overqualified. I know my explanation left a lot out and may have seemed overly simplistic, but many newer traders will not be at all familiar with the academic "inside baseball" controversies on probability and how it relates to trading. My purpose as I said was mainly to get some folks from "trading math 101" into "trading math 202."
  • #10
  • Dec 14, 2007 12:57am Dec 14, 2007 12:57am
Quoting StockKJay
Hello Capitalist,

Even this scenario, as simple as it sounds, would require fairly unusual and sought after circumstances.

Allow me to try and explain my position. One would have to be trading in the middle of a range for this to be in effect. A pair would need to be stuck in that range, have absolutely no measurable pattern of how it behaves in that range and have no external or changed momentum characteristics. Then, if a trade were taken in the middle, an equal TP and SL would more than likely bring the % to within +/- 1. One would assume you could find these conditions if searching hard enough.
Hi StockKJay,

If I understand your position correctly you are saying that different current market conditions will have different distributions of future price movement, and I couldn't agree more. A market that has a steady tight trend or which is influenced by strong fundamental factors will certainly have a higher probability of moving by X pips in one direction before moving X pips in the other. The key is identifying those conditions and knowing how they skew the probabilities. That's an edge.

But many newer traders may not know whether such a market is full of momentum and likely to keep going for another six months or whether it is overbought and prone to a strong correction tomorrow. Heck, even veteran trades have a hard time with that. But especially for a newer trader who might not even consider the question, that trader's entries might as well be random. Even if the market is not random, sometimes traders are.


Quote
**Question**

Having stated that, how would you examine the statistics of the previously mentioned trending market? ...

...This is something I am currently exploring, though I admit I am having a tough time finding a formula to use. I would very much like to be able to factor the win % of a set TP with an equal SL in a given market condition.

This is something that I've been exploring through private research for years. I don't know about finding a specific "formula" but the basic methodology I use is to create a way to rigorously define some set of market conditions and then use historical market data to explore how the market subsequently reacted to instances of each different condition. Then there's the question of whether what I've found is a real influence on price or just a statistical artifact, so you get into issues like datamining, out-of-sample testing etc. which I won't bore anyone with here. So I guess the basic answer to your question is just "historical statistical research."
  • #11
  • Dec 14, 2007 1:15am Dec 14, 2007 1:15am
Quoting Plutonite
Jay is correct: the 50% hit rate given 1:1 risk/reward is not true.

It would only be true if the market was a random walk in the space of possible prices but the market is not a random walk model. This has been mathematically proved since the 90's in various rigorous studies in the Journal of Finance..etc. The market is only rarely random, especially when there is no speculative activity. Your probability of success is a function of your risk/reward, yes, but there are also other unknown factors. Just because they are unknown does not mean they are not there.

Hi Plutonite!

I certainly don't want to give anyone the impression that I'm a Random Walker! I won't put any links here, but if you google "A Random Rant on the Random Walk" and "Beyond the Random Walk - A New Market Paradigm" you will find some of my articles on this. In the first one I have fun giving Burton Malkiel a piece of my mind.

However, I can see how my original post might have been construed that way. Hopefully if you've read my last few posts you'll see that I started the thread more for an audience that is somewhat less familiar with the more advanced issues. You are sort of like a physicist in a first year algebra class. :

Quote
Using previous information (e.g patterns, order flow information) you can have at least some of those factors on your side.

I couldn't agree more, and the point of my original post was how important it is to find such an edge. Without such an edge though, the new trader falls victim to the "if I just set my TP at 3x my SL I will be rich" trap. My goal was to point out the fallacy in the view that a simple MM approach is all that's required for success. It isn't. As you and other posters here have pointed out, a trader needs an edge.

Thanks to you and everyone else for their contributions so far!
  • #12
  • Dec 14, 2007 1:18am Dec 14, 2007 1:18am
Whoa. My fingers are tired! :surprised I'm heading off to bed...
  • #13
  • Dec 14, 2007 6:51am Dec 14, 2007 6:51am
Quoting capitalist88
Actually, without some definite edge, expectancy is always negative because of spreads and commissions. Here's why.

Most traders will be familiar with the standard expectancy equation:

W = win rate
G = average Gain (or predefined TP if there is one)
L = average Loss (or predefined SL point if there is one)

Expectancy = WG - (1-W)L

This is like trading math 101 but it is very compelling to new traders who think that it is the key to sure consistent success. "After all," they think, "if I just cut my losses to 10 pips, and let my trades run to 30 pips, then even if I only win 30% of the time then I would still be ahead!"

That is of course because (0.30 x 30) - (0.70 x 10) = 9 - 7 = 2 pips/trade.

Wow! Any dolt can be right only 30% of the time, so all I have to do is make a million trades this year and I'll be up 2 million pips! Right? Right????

Ummm, no. :

Let's get beyond trading math 101 to the next step. What's wrong with the train of thought above is that the win rate is actually a function of your risk/reward ratio.

Suppose you set your stop at 1 pip and your TP at 800 pips on every trade. How many do you think you'd win? Your win rate would be virtually zero. Now if you reversed the situation and set your TP at 1 pip and your stop at 800 pips then your win rate would be very close to 100% If your SL and TP are at the same distance from your entry, then your win rate should be about 50%. In fact, your win rate (all else being equal) should be equal to the distance to the stop divided by the total distance from the stop to the target.

For example if your SL is 20 pips from entry and your TP is 80 pips from entry, then your win rate on trades like that should be around 20%. If you set a stop at 20 pips and a TP at 5 pips then your win rate should be about
20 / (20 + 5) = 20/25 = 80%

So I'm saying that in terms of the variables in the standard "expectancy" equation, the win rate is actually equal to: L / (L+G)

So let's plug that into the expectancy equation and see what we get shall we? We started with:

E = WG - (1-W)L

but now we know that W = L / (L+G) so...

E = LG / (L+G) - L + L^2 / (L+G) note: L^ is "L squared."

multiplying the middle term on the right by (L+G)/(L+G) which is just 1:

E = LG / (L+G) - L(L+G) / (L+G) + L^2 / (L+G)

combining into one fraction with a common denominator:

E = (LG - L(L+G) + L^2) / (L+G)

E = (LG - L^2 - LG + L^2) / (L+G)

E= 0 / (L+G)

E = 0


And that in a nutshell is why trading can be difficult. Just depending on old pat bromides like "cut your losses short and let your profits run," or "you can't go broke taking profits," although they sound good will not really get you anywhere.

Sure, if you cut your losses short, you know what you get? A lot of little losses which just about equal your very infrequent big wins. If you follow the opposite advice that "you can't go broke taking profits," and decide to use wide stops and tight targets guess what? You get a lot of little profits that just about make up for that huge loss you take every so often.

All of these cute slogans do nothing more than represent different points along the risk/reward spectrum, but traders should be aware that each of those points has an expectancy of about zero. And that's not including costs. So forget the cute slogans. The key to trading successfully is to find an edge that moves that E value slightly to the positive side of your costs.

Very interesting, however, you are looking at the whole thing from just 1 angle...The risk/reward ratio is NOT related to the Win/Loss ratio the way you stated....

What you are saying is that while the win/loss ratio moves from 0 to 100%, your risk/reward is actually moving the other way and thus always having a near zero expectancy...

This is true for the majority of trading systems out there...Millions of trading systems have around zero expectancy...However, very few strategies manage to tweak this to produce high positive expectancy...Remember that your risk/reward ratio is just a theoritical thing...If you are using a trend following strategy with a risk/reward of 3:1, that's risk is 3 times the reward, you are not actually going to close out your trade at 1 reward or profit, you will rather ride profits once the trade goes in your favor....Trade management is capable of enhancing the THEORITICAL risk/reward ratio of your strategy....The 1 time you catch a big, really big trend will produce a trade of a 1:10 risk/reward trade in reality versus an original theoritical 3:1 risk/reward trade....


Thanks,

Nader
  • #14
  • Dec 14, 2007 5:16pm Dec 14, 2007 5:16pm
Quoting narafa
Very interesting, however, you are looking at the whole thing from just 1 angle...The risk/reward ratio is NOT related to the Win/Loss ratio the way you stated....

What you are saying is that while the win/loss ratio moves from 0 to 100%, your risk/reward is actually moving the other way and thus always having a near zero expectancy...

This is true for the majority of trading systems out there...Millions of trading systems have around zero expectancy...However, very few strategies manage to tweak this to produce high positive expectancy...Remember that your risk/reward ratio is just a theoritical thing...If you are using a trend following strategy with a risk/reward of 3:1, that's risk is 3 times the reward, you are not actually going to close out your trade at 1 reward or profit, you will rather ride profits once the trade goes in your favor....Trade management is capable of enhancing the THEORITICAL risk/reward ratio of your strategy....The 1 time you catch a big, really big trend will produce a trade of a 1:10 risk/reward trade in reality versus an original theoritical 3:1 risk/reward trade....


Thanks,

Nader
Hi Nader,

You're saying that there are ways (good system, good methods of trade management, etc.) to get the expectancy above zero. That's what I'm saying too. Expectancy is zero unless you have an edge.

But if a new trader fails to do those things, doesn't do the work to find an edge and just relies on the idea that a simple MM rule will ensure success, then they're going to be in trouble.

As you said, this is true for the majority of trading systems out there. I don't know if you read the rest of the thread, but it was to exactly that majority that I addressed the original post. More experienced traders like yourself already know these things, but newer ones may not. So I presented a basic example to show what the issue is. I agree completely with you that there are ways to get expectancy above zero. If I didn't I wouldn't bother trading.
  • #15
  • Dec 14, 2007 5:54pm Dec 14, 2007 5:54pm
Nice thread Cap' , definitely food for thought.
Personally, I abore stops and tp's, I do believe in 'emergency' stops though, just in case the lights go out you don't want your book to dissolve without your help at least

I will set them if I leave a trade open and I can't monitor it though but that's another story.

What you are saying (I think) is that there's no benifit in sticking to a set R/R ratio and I totally agree, nothing is certain and by doing this you are 'predicting' or even worse 'hoping' the outcome will be in your favour which is dangerous play.

What's the answer?
I don't know, but every trader has their own risk tollerance and strategy (if they don't, they need to find it quick) and knowing about what you have put forward is at least something to consider in the whole trading picture.

Thanks Cap'
Mick.
  • #16
  • Dec 14, 2007 6:12pm Dec 14, 2007 6:12pm
My original post was designed to help newer traders think more deeply about the relationship between money management rules and their win/loss ratio and how they can't affect one without affecting the other.

However I see that most of the posters here are more knowledgeable about the subject, so now I'd like to take it up a notch by suggesting a simple experiment that any trader can do.

There's been a common theme among some of the responses that the relationship I talked about between risk/reward and win ratio isn't really very strong. The basic objections seem to be of the following types:

1. Market conditions change and under each different market condition (trending, ranging, volatile, etc.) the relationship between risk/reward and win ratio will be different too.

2. Some traders just rock and will have a good edge, good trade management skills, etc. and this will cause the relationship to break down to some degree.

3. There is a theoretical advantage to money management because of closed ended risk and open ended reward potential ("black swan", "fat tails" inevitability of rare occurrences, or whatever you want to call this concept).

Now I agree in principle with each of those objections, but I also know the potential pitfalls of relying to heavily on each of them as well. I won't adress those individually in this post, but I would like to put my underlying assumption of a strong relationship between risk/reward and win ratio to the test. I propose that anyone reading this do the following experiment and report their results back to this thread (I have done this test with my own trades in the past):

What you do is simply plot a moving average of your Actual Win Rate (AWR) and your Required Win Rate (RWR) over time and note how the lines tend to hug each other. If you already have a log of your last 100 trades or whatever then try taking a 10-trade moving average (or whatever you think is a reasonable amount) over this whole period. If you don't have a record of your trades then this is a good time to start. :

Assuming you're using a 10 trade moving average, here are what the values mean:

Actual Win Rate (WR) is just the number of trades with a profit divided by 10. So if you had 4 winning trades over the last 10, then your WR = 40%. Everyone's familiar with this one of course.

Required Win Rate (RWR) is the win rate that you would need to have in order to be at breakeven or better given your average gain and average loss. RWR = L / (L + G) where L is average loss on losing trades and G is average gain on winning trades. For example if my four winning trades had an average gain of 30 pips and my six losing trades had an average loss of 10 pips, then my Required Win Rate is 10/(10+30) = 25%. This means that given my current trading behavior (keeping losses at about 1/3 of gains) I would need to have a win rate of at least 25% to be profitable.

This is pretty easy to put together in Excel, so I set up some made up data for 20 trades to show the process. There is a chart below the table showing the Actual vs. the Required Win Rate lines.

Attached Image (click to enlarge)
Click to Enlarge Name: example.JPG Size: 63 KB


I think that if you plot your trading results out this way, you will be surprised at how closely the two lines hug each other. From a mathematical perspective (which is just one of many perspectives), trading is all about keeping that blue line above that purple line consistently.

And now, it's time for pizza....

  • #17
  • Dec 14, 2007 6:19pm Dec 14, 2007 6:19pm
Quoting MickD
Nice thread Cap' , definitely food for thought.
Personally, I abore stops and tp's, I do believe in 'emergency' stops though, just in case the lights go out you don't want your book to dissolve without your help at least

I will set them if I leave a trade open and I can't monitor it though but that's another story.

What you are saying (I think) is that there's no benifit in sticking to a set R/R ratio and I totally agree, nothing is certain and by doing this you are 'predicting' or even worse 'hoping' the outcome will be in your favour which is dangerous play.

What's the answer?
I don't know, but every trader has their own risk tollerance and strategy (if they don't, they need to find it quick) and knowing about what you have put forward is at least something to consider in the whole trading picture.

Thanks Cap'
Mick.

Hi Mick!

I see you all over this place, LOL.

I am actually following your lead in a way as I am moving more toward setting emergency stops far away and exiting on price action signals rather than using a stop as an exit strategy. I also have pretty much abandoned TPs because of the "open ended reward" argument.

However, even for traders like us who don't use set stops and TPs, you will still have an average gain and average loss on your trades. So even if you don't use the actual orders, it's AS IF you do, just because of whatever trading behavior you have. Take a look at the post before this one for an example of how I track trading behavior in terms of risk/reward vs. win rate.
  • #18
  • Dec 14, 2007 6:30pm Dec 14, 2007 6:30pm
I started trading with only a little bit of TA knowledge and lost quickly. I tried everything I could possible think but every method worked out the same. After 2 years I was down $20,000 and thought I was the worst trader ever! but get this, I did over 1000 trades in that two years, I managed to do $80,000 dollars worth of trades with a $20,000 initial stake. Heres the catch the commision per trade was $20 which equalled $20,000 in spread. I actually broke even! my losses were due to spread. the expectancy played out to almost exactly zero except my broker had a slight edge and cleaned me out with it.
BTC Return This Year: na
  • #19
  • Dec 14, 2007 7:00pm Dec 14, 2007 7:00pm
Quoting M.A.C.Doug
I started trading with only a little bit of TA knowledge and lost quickly. I tried everything I could possible think but every method worked out the same. After 2 years I was down $20,000 and thought I was the worst trader ever! but get this, I did over 1000 trades in that two years, I managed to do $80,000 dollars worth of trades with a $20,000 initial stake. Heres the catch the commision per trade was $20 which equalled $20,000 in spread. I actually broke even! my losses were due to spread. the expectancy played out to almost exactly zero except my broker had a slight edge and cleaned me out with it.
Nice story M.A.C.Doug! that's a pretty good example of what Cap' is talking about.
I guess the next step is how do we tip the scales in our favour??
That will be the 'edge' we need, any thoughts??
  • #20
  • Dec 14, 2007 7:52pm Dec 14, 2007 7:52pm
Quoting MickD
Nice story M.A.C.Doug! that's a pretty good example of what Cap' is talking about.
I guess the next step is how do we tip the scales in our favour??
That will be the 'edge' we need, any thoughts??
One way is to trade less often on longer time frames. That is one way the spread can become less significant.
0 traders viewing now