Long term Profits within Negative Expectancy | Page 15 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

What is Expectancy? 13 replies

System Expectancy - Monte Carlo Simulation 5 replies

Short Term / Medium Term / Long Term? 3 replies

Random Market vs Set Expectancy 8 replies

Expectancy is always zero 41 replies

Attachments: Long term Profits within Negative Expectancy
Exit Attachments

Long term Profits within Negative Expectancy

  • #281
  • Aug 25, 2009 8:35am Aug 25, 2009 8:35am
DT,

I don't think we have to go that far, initally. Let's try this:

1) Every trade begins life with negative ticket balance--- we must pay the spread before we play.
2) The market price is, at any given moment, what the market says it is.
3) You place a trade with some knowledge you believe the market does not have or does not recognize. If the market had digested that knowledge and your assumptions were correct, the price would not be where it presently is.
4) What made you sure enough about that knowledge that it caused you to take the trade? Let's analyize that and see if it holds up to logical reasoning. If so, we can proceed with testing and attempt to measure expectancy. I am afraid that most traders would fall apart at this step because they could not give a clear logical reason for entering a trade.

For example: When price approaches a "00" number, it will bounce at least once off that number and retrace at least 10 pips. Not everytime, of course, but enough times to be statistically significant. Now that is a perfectly good market assumption that would be worthy of further testing. As it turns out, this assumption would be wrong and you would get your lunch eaten if you employed it. As you would with every other market assumption I have tested (except for a couple I am still working with). I do admit, however, that my work is not exhaustive. There may well be some methods that I am not aware of that work... hell, I hope there are or all my work is in vain. Alot of posters seem to think that just becuause their equity balance is positive at the moment then all of the work has been done and they are on their way to a life of freedom as a daytrader. This is dangerous thinking and has hurt a lot of people. I hope my little thread gives a few of them a reason to analyize exactly WHY their balance is positive and if those reasons give them hope to stay that way.

Regards
  • #282
  • Edited 10:01am Aug 25, 2009 8:56am | Edited 10:01am
Quoting FX-Petra
3) You place a trade with some knowledge you believe the market does not have or does not recognize. If the market had digested that knowledge and your assumptions were correct, the price would not be where it presently is.

Which came first, the chicken or the egg?

At some point the market will have had to start digesting that knowledge/information, the idea is to realize when that digestion process begins and in which way it is being digested, and then try to follow by getting in on the action at strategic points. That can be by either using technical analysis or by using fundamentals or by using a combination of both (my preference).

Add to that the 'natural' ebb and flow of the market as traders open/close positions sometimes at significant levels, obviously that isn't priced-in knowledge, and price moves accordingly (move/retracement/expansion).

Quoting FX-Petra
4) What made you sure enough about that knowledge that it caused you to take the trade?
Taking into account all the factors (charts, fundamentals, whatever), on the balance of probability price had more reason to go one way than the other in my opinion (the path of least resistance perhaps). The rest is math, risk:reward. Whether win, lose, or draw for that particular trade it's irrelevant, historical statistics show that overall the account will show positive results. As you pointed out historical results are no guarantee of future performance but what else do we have?

I'm sure this is all stuff you know already.
  • #283
  • Aug 25, 2009 10:13am Aug 25, 2009 10:13am
Quoting ThePipDoctor
I think he has answered it in a way, and what hes saying is intuition plays an important part. And remember, just because you have a postive expectancy doesnt mean you can print money at will, granted with decent capital you can make a lot but if you want to stay in the game, you need to keep a sensible risk level...

Come on Petra, I know you are firm in your beliefs because of your past failures, but you do need to keep an open mind. You've admitted your testing wasn't completely comprehensive.

Can I ask for the purposes of discussion, since...
Negative expectancy fluctuates (as a percentage) depending upon your trading style trading style. I will use 1-1 R/R ratio for simplicty's sake. Please do not report that if you use a diffrent R/R that the edge will change. It will not. Example:

Enter a trade, spread is 3 pips. 15 pip stop loss and 15 pip profit. Your expectancy is--- 12/30=40.0% or -10.0%. Use the same trade but change your stop loss and TP to 50. Your expectancy is 47/100= 47% or -3%. It is well established that larger your SL and TP, the better you odds become. You are giving market volitilty a greater chance to work in your favor. Problem is, even if you go to a 1000 pip stop loss and 1000 pip profit the negative aspect of your expectancy never goes away. I will admit that when you have large stop losses and TP's the negative expectancy starts getting pretty small, and you would only need a tiny amount to overcome it and become positive. I believe that is why LONG TERM traders have the most success. They can use long term underlying fundamentals of a currency to gain an edge. Daytraders are playing a diffrent game. Using a 15 pip S/L and T/P, I do not believe there is anything presently availble to overcome -10% expectancy. I hope there is, and wish there were, and pray there might be... but zilch so far.

Regards
  • #284
  • Aug 25, 2009 10:41am Aug 25, 2009 10:41am
[



Taking into account all the factors (charts, fundamentals, whatever), on the balance of probability price had more reason to go one way than the other in my opinion (the path of least resistance perhaps). The rest is math, risk:reward. Whether win, lose, or draw for that particular trade it's irrelevant, historical statistics show that overall the account will show positive results. As you pointed out historical results are no guarantee of future performance but what else do we have?

I'm sure this is all stuff you know already.[/quote]


PM,

I appreciate your honesty. Few here are so brave. You trade your opinon. Nothing wrong with that. It is your money. I'm sure you are a very astute trader and strive to make the best decisions possible. I too have tried to trade in this manner. I have had a streak of great success, making thousands and thousands of dollars. But things reversed. I had to come to grips (with the help of a friend) that I was engaging in a negative expectancy activity and reversal was just a matter of time. I then started to find a way to gain an edge. I have come close with a couple of systems... but still no success.

On a positive note, I will give you this. It IS possible to win at this game, even with Neg. Exp. Here's how:
Pick a dollar figure that you want to achieve. Once you achieve this figure, walk away. Insane profits are possible in the short term. The odds, while never in your favor, can still give you a reasonable chance of success given adquate capital and astue money management. Take that 50K equity and run it to 500K and then get out. It has been done before and will be done again.

Regards
  • #285
  • Aug 25, 2009 11:32am Aug 25, 2009 11:32am
FX-Petra,
I have no doubt, that your math is correct, but market is not a mathematical model.
At its core it is wild herd of madmen driven by greed and fear.
My suggestion would be, that secret to cracking this baby would be in crowd psychology ( sociology).

Regards,
AT
  • #286
  • Aug 25, 2009 11:50am Aug 25, 2009 11:50am
Quoting ALThau
FX-Petra,
I have no doubt, that your math is correct, but market is not a mathematical model.
At its core it is wild herd of madmen driven by greed and fear.
My suggestion would be, that secret to cracking this baby would be in crowd psychology ( sociology).

Regards,
AT
ALThau,

As I stated from my very first post, I realize it is difficult for most people to accept mathematics. This is almost universally true when mathematical outcomes conflict with our beliefs about what is and is not true about our world. All most all posters have come down into 3 categories:

1) My math is wrong (calculations)
2) I am impoperly applying math to the market from the standpoint of methodology.
3) Math is not big enough to 'capture' the market.

It seems you are in group # 3.

My responses are:

1) If so-- please point it out.
2) If so-- please point it out. Your opinions or feelings don't count. Please show me with hard data or, in the alternative, logical reasoning.
3) This cannot be proven or disproven. Math has served as the building block for all nature and the history of man. Circumnavigation of the globe, travel to the moon, and splitting the atom just to name a few. If you believe that the forex market is to complex to be captured by mathematics, then we are at an impass. I wish you all the best....

Regards
  • #287
  • Aug 25, 2009 11:50am Aug 25, 2009 11:50am
Quoting FX-Petra
Pick a dollar figure that you want to achieve. Once you achieve this figure, walk away. Insane profits are possible in the short term. The odds, while never in your favor, can still give you a reasonable chance of success given adquate capital and astue money management. Take that 50K equity and run it to 500K and then get out. It has been done before and will be done again.
There's no need to go after insane profits though, just a decent and consistent return, and there's nothing to change that as far as I can see. If an edge starts to become less effective then it's going to become obvious fairly quickly, that's the time to stop and review in my opinion, why do it before?

As the time limit of profitability hasn't been defined how do we know we'll still be trading when it runs out, we may be retired or worse, dead!
  • #288
  • Aug 25, 2009 11:54am Aug 25, 2009 11:54am
Quoting FX-Petra
DT,

I don't think we have to go that far, initally. Let's try this:

1) Every trade begins life with negative ticket balance--- we must pay the spread before we play.
2) The market price is, at any given moment, what the market says it is.
3) You place a trade with some knowledge you believe the market does not have or does not recognize. If the market had digested that knowledge and your assumptions were correct, the price would not be where it presently is.
4) What made you sure enough about that knowledge that it caused you to take the trade?...
Petra,

Initially, all the edges which we as traders carve out, are based on observation; like you believe that your observation of the round number occurrence has enough statistical relevance to give it further thought.

The charting medium has given traders more and more opportunities to search and find such occurrences. They can be build on many different aspects of price behaviour but all of them should have a firm grounding in the 'time' aspect in terms of the traders managability. What I mean is that many or most of the systems displayed here and everywhere else are indicator laden and signals pup up like mushrooms in such a way that the regular trader would have to stay in front of the screen 24H in order to catch all the significant signals.

So, in my view, a fundamental point (unless you leave all trading to the EA's), your edge will have to be created in such a way that signals will happen in your time zone. What good is backtesting if you can't follow the model in real-time.

Will finish this later.

regards
daytrading
Enter Signature
  • #289
  • Aug 25, 2009 12:35pm Aug 25, 2009 12:35pm
"If you believe that the forex market is to complex to be captured by mathematics, then we are at an impass. ..."


Nah, actually, I believe mathematics are descriptive and predictive.
It's called statistics.
And, I do believe you can apply statistics to human behavior.
What I don't believe, is a possibility of creating mechanical system for beating this game in changing environment.

Regards
  • #290
  • Edited 1:16pm Aug 25, 2009 1:00pm | Edited 1:16pm
Quoting FX-Petra


3) This cannot be proven or disproven. Math has served as the building block for all nature and the history of man. Circumnavigation of the globe, travel to the moon, and splitting the atom just to name a few. If you believe that the forex market is to complex to be captured by mathematics, then we are at an impass. I wish you all the best....

Regards
FX-Petra,

Today's mathematical tools in and of and by themselves are most definitely powerful enough to capture a complex system such as forex markets. However what is at the core of this problem is:

1) modeling such a complex environment as a map for the artificial processor
2) modeling the behavior of the actor (trader) for automation purposes (in case of
backtesting) to be mimicked by the artificial processor
3) and finally the implementation of the interaction of the 1) and 2) for a complete
model (i.e. sythetic testing environment with no feedback)

and that's where the impasse occurs; whereas in engineering and scientific applications you named, many of feedback behaviors of the components involved can be successfully simulated (e.g. the lack of gravitational force, just to name one).
Hence the unsatisfactory backtest results you are talking about.
Maybe this could be a worthwhile project to ponder.

regards,
GM
Hear the music before the song is over ...
  • #291
  • Aug 25, 2009 1:25pm Aug 25, 2009 1:25pm
Quoting FX-Petra
All most all posters have come down into 3 categories:

1)...
The way I see it, the problem is that you try a priori to enforce a set of logical, mathematical rules on an extremely complex system - a system for which their exist no accepted standard models, contrary to the relatively simple systems you mention in 3).

Markets and their mechanics are not necessarily "logic" in the sense they can be "figured out"/modelled a priori. So the problem with 2) is that

a) Logical reasoning is completely insufficient.
b) No one is going to point it out with hard data as it would expose their edge for no reason (and by the way, NDA's are complete bullshit in this business).

I realize that I am not really adding anything to the discussion as this has already been mentioned in one way or another. I can only say that, interestingly enough, the best (auto)trades I know all have a background in applied physics/computer science. Non of them are from pure math...
  • #292
  • Aug 25, 2009 1:28pm Aug 25, 2009 1:28pm
I'll say this for you guys.... Keeping up with all the ideas and views presented in this thread is a dizzying process. I appreciate all the posts. It is rare that a discussion on so irritating a topic could hold up this long with continued new insite and civil discourse.
  • #293
  • Aug 25, 2009 1:34pm Aug 25, 2009 1:34pm
Quoting FX-Petra
I'll say this for you guys.... Keeping up with all the ideas and views presented in this thread is a dizzying process. I appreciate all the posts. It is rare that a discussion on so irritating a topic could hold up this long with continued new insite and civil discourse.
It's rare that a discussion like this has continued for so long with such opposing views without descending into the usual FF tat. Credit goes to the contributers.

I'm just lurking and enjoying an excellent thread.
  • #294
  • Aug 25, 2009 2:46pm Aug 25, 2009 2:46pm
Quoting daytrading
Petra,

Initially, all the edges which we as traders carve out, are based on observation; like you believe that your observation of the round number occurrence has enough statistical relevance to give it further thought.

The charting medium has given traders more and more opportunities to search and find such occurrences. They can be build on many different aspects of price behaviour but all of them should have a firm grounding in the 'time' aspect in terms of the traders managability. What I mean is that many or most of the systems displayed...

Ok, carrying on....

Another aspect in your structure to approach maybe a little different is what you refer to as 'negative' expectancy. It's a business and like in every good business, you have to enquire about the profit margin. That means you cannot apply a model that has the law of large numbers as a foundation with a 1:1 risk/reward ratio - of course it will be resulting in negative expectancy.

Here is a thinking exercise to get the head 'round it: if you contemplate about a model that gives you (or should give you) 10 pips profit for a 10 pip stop and you are content with the fact that your model presents you with these opportunities, then why not ask for 15 pips on the win versus 10 pips on the stop side? So, by setting your risk/reward ratio from the outset to 1.5:1, you will never have to worry about the 2 pip spread on eur/usd anymore; costs are covered.

If you are at the stage that you believe your edge will deliver 10 pips, why not 15 pips? If you say you can't find the cofidence that the market will move 15 pips in your favour, you won't have the confidence for 10 pips either - making trading a futile exercise as a whole.

There are models out there which have negative risk/reward ratios (20 pip stop vs. 10 pip gain), which are balanced out by 80% win/loss ratios (mostly algorithmic, placing 500 trades/day).

If your model is anchored in the 50/50 win loss bracket, you need to construct your edge in a way that costs become absorbed by a higher risk/reward ratio - otherwise, your theory will prove correct and you ultimately lose over time.

regards
daytrading
Enter Signature
  • #295
  • Aug 25, 2009 3:01pm Aug 25, 2009 3:01pm
Quoting daytrading
if you contemplate about a model that gives you (or should give you) 10 pips profit for a 10 pip stop and you are content with the fact that your model presents you with these opportunities,
DT, this hasn't been my experience at least in the discovery of an edge, only in the application of one. Now, out of thousands of systems, I would only classify three of mine having any worth, and they all share the same characteristic. You can trade them 1:1 profitably, although more optimal (curve-fitted?) gains are possible at different r:r's.
I share the same view as Petra on this - I can't see that the moving of SLs/TPs on a system would discover an edge that would never otherwise be found (unless of course your strat is based on stops).
  • #296
  • Aug 25, 2009 3:08pm Aug 25, 2009 3:08pm
Quoting daytrading
Ok, carrying on....

Another aspect in your structure to approach maybe a little different is what you refer to as 'negative' expectancy. It's a business and like in every good business, you have to enquire about the profit margin. That means you cannot apply a model that has the law of large numbers as a foundation with a 1:1 risk/reward ratio - of course it will be resulting in negative expectancy.

Here is a thinking exercise to get the head 'round it: if you contemplate about a model that gives you (or should give you) 10 pips profit for...
I don't disagree with what you have said. If you have a 50% win ratio with a 1/1.5 RR, you are in the money. However, I contened that if you move your RR to 1/1.5, your win % will drop to 8/25=32% (based upon the 15 pip win or 10 pip loss). Don't get me wrong... I hope you are right. If there is a system that yeils a 50% win ratio with 1:1.5 risk/reward it would change my life. You don't even have to explain the system to me. I respect your thoughts and just your assurance that such a system is workable over the long term is enough to re-energize me and give hope.

Regards
  • #297
  • Aug 25, 2009 3:13pm Aug 25, 2009 3:13pm
Quoting triphop
DT, this hasn't been my experience at least in the discovery of an edge, only in the application of one. Now, out of thousands of systems, I would only classify three of mine having any worth, and they all share the same characteristic. You can trade them 1:1 profitably, although more optimal (curve-fitted?) gains are possible at different r:r's.
I share the same view as Petra on this - I can't see that the moving of SLs/TPs on a system would discover an edge that would never otherwise be found (unless of course your strat is based on stops).
Maybe I didn't make this point too clear: you don't gain the edge from attempting to increase the R/R by itself. The edge is one thing (for example finding a phenomenon that yields movement in your desired direction - however many pips). Dealing with the costs (of the spread) is another.

If your model is profitable on a 1:1 basis, it MUST win more than 50% of the time - otherwise over the long term (as per Petra's main point), you lose.

If your model wins exactly 50% of the time, with 10pip target and 10pip stop, your R/R is in fact 1:1.2 if you close your position if it hits -10 (since it took the price to move only 8 pips (assuming 2 pip spread) before reaching your stop. Don't forget by the time you enter, you are 2 pips closer to your stop - but still 10 pips away from your target.

regards
daytrading
Enter Signature
  • #298
  • Aug 25, 2009 3:38pm Aug 25, 2009 3:38pm
Quoting FX-Petra
I don't disagree with what you have said. If you have a 50% win ratio with a 1/1.5 RR, you are in the money. However, I contened that if you move your RR to 1/1.5, your win % will drop to 8/25=32% (based upon the 15 pip win or 10 pip loss). Don't get me wrong... I hope you are right. If there is a system that yeils a 50% win ratio with 1:1.5 risk/reward it would change my life. You don't even have to explain the system to me. I respect your thoughts and just your assurance that such a system is workable over the long term is enough to re-energize...
Petra,

Since this is not a coin flip or a roulette wheel, winning X does not have to equate to losing X on a 50/50 basis. I give you an example:

One of my models has a pretty balanced win/loss ratio - i.e. over the course of a month, quarter and year, it comes to more or less 50% winning and 50% losing trades (1200 on avg per year). The model relies on a less frequent but reoccurring phenomenon which delivers days (not indiv. trades) with a 1:4 risk/reward ratio. Since I never really know when these days happen in advance, I had to create a way to stay almost constantly in the market (except overnight, which still makes it an intraday model) with a balanced risk - even ever so slightly tilted against me with the spread - like your theory.

The exceptional days take care of the negative skew.

regards
daytrading
Enter Signature
  • #299
  • Aug 25, 2009 3:41pm Aug 25, 2009 3:41pm
Quoting daytrading
Hi Hanover and Global Macro and Chicky,

Great posts. I think we might be getting somewhere here.

................

P.S. ....it has become interesting again.
Hi Daytrading,

I'm enjoying your insightful take and contributions.
One of the biggest obstacles to success in short-term especially intraday trading is the constraint that companies put on their traders regarding the time-in-the-market, drawdown provisions, and frequency of trading whereas macro-oriented traders do not suffer from these shortcomings and can more extensively complete the trading strategy cycle without unnatural imposition of those restrictions. What's bothering me and frankly beyond logic to me is what's the benefit for the owners of these companies to enforce such inflexible rules.

Again please keep up your very useful contributions.

regards,
GM
Hear the music before the song is over ...
  • #300
  • Aug 25, 2009 3:50pm Aug 25, 2009 3:50pm
Quoting GlobalMacro
Hi Daytrading,

I'm enjoying your insightful take and contributions.
One of the biggest obstacles to success in short-term especially intraday trading is the constraint that companies put on their traders regarding the time-in-the-market, drawdown provisions, and frequency of trading whereas macro-oriented traders do not suffer from these shortcomings and can more extensively complete the trading strategy cycle without unnatural imposition of those restrictions. What's bothering me and frankly beyond logic to me is what's the benefit for the owners...
Hi GM,

Thanks for the compliments.

I can only comment from personal experience which might well differ from that of other proprietary traders.

The founders of the firm I work with have floor trader roots (LIFFE) and taking position overnight was never an option for this breed of trader.

My assumption is that this has translated into the type of restriction you've mentioned. We have a fund incorporated in the company that works on a macro basis and does not underly such constraints but is in turn not accessible for the daytraders.

Freedom is essential in my view.

regards
daytrading
Enter Signature
1 trader viewing now