I like the way you think.At best, it seems a trend could be more accurately described as the prevailing market sentiment. Who's to say if and when this sentiment will change. Some of the best advice I've seen mentions ways to read what the markets true intensions are. A mixture of Fundamentals and Technicals that point to an imbalance that will need to be corrected in the future. Unfortunately for small guys like me, this tends to be very long term advise which requires a very large stop.
Much of What You Know About 4X is Wrong!
- #62
- Edited 11:35am Oct 21, 2007 11:21am | Edited 11:35am
- | Joined May 2007 | Status: Seasoned Trader | 820 Posts
Oh oh! The cryptology department has invaded MY thread.
Please speak directly or pass out decoders to everyone else. This is a discussion thread in a discussion forum and till now all was going very smooth. Secret hand shakes are better left elsewhere.
Ok now try not to scratch each others eyes out.
Xyrus, welcome to the forums. As a nuclear engineer (former) I enjoyed the physics reviews, especially your inference to the Heisenberg Uncertainty Principle. As a trader (present) I think you overlooked one of the most basic laws of physics. Namely, Newton's first law.
My only other point of contention is to point out that markets are driven by human emotion and a herd mentality whereas the laws of physics are not. Electrons do not make quantum jumps to a higher orbit when they get scared.
Emotion and herd mentality are indeed factors in the market. However, even herd dynamics have a mathematical model. If the "alpha males" of a herd believe there is a threat, then the rest of the herd will follow. If enough of the "regulars" believe there is a threat, then the herd will also move.
Or in short, a trend is a self-fulfilling prophecy. If a critical mass of traders believe there is a trend, then a trend will form. Or in physics, if enough particles move in a particular direction, it will move the rest of particles with it.
And phil is correct, electrons do not get scared. But they can get excited
.My point about Soros is not how he trades now. Obviously the market looks very different when you're sitting on top of billions of dollars. However, he did not get his billions day-trading the stock market or Forex. My point was that coming into the Forex or other market expecting to daytrade your way to millions is a recipe for disaster. Making a 30% ROI is pretty damn good and there aren't many pro's who can say they can do that for any length of time. "Good investing is boring," as Soros says.
Cryptography department?
Well, a lot of the technical indicators do get their basis from signals analysis. In fact, the standard exponential moving average is just a bandpass filter.
~X~
- #65
- Edited 3:22pm Oct 21, 2007 3:12pm | Edited 3:22pm
- Joined Sep 2006 | Status: ... | 8,112 Posts
Once you have devised a trading method that works for you - trend retracement and trend reversal and trend continuation technical indicators sorted, system strategy back tested and paper traded successfuly, scope of achievable pips determined, stop loss strategy planned, live trading commenced with suitable trading bank and prudent money management -
Soon after - YOU LOSE THE PLOT!!!!!
Human nature is the major failing of unsuccessful traders...
A Blackjack pro will always double down whenever he holds 11 versus the dealer's 6. Even if he lost the last ten times doing so, he will continue to make the mathematically correct play, because he can be certain that - no matter what has happened in the recent past - the odds are significantly in his favor upon playing that way.
So why is it so much more difficult to "keep to one's rules" when trading? Is it because, whereas the Blackjack pro knows (given 11 vs 6) that the odds are in his favor, that the trader - no matter how extensive his backtesting, or how exhaustive his sample size - knows that probabilities can't be precisely calculated, and is therefore constantly haunted by the idea that his test results may be statistically uncertain? Or is it because, whereas a Blackjack hand is completed almost instantly, leaving little time to panic or agonize, the trader becomes spooked watching price dance up and down, favorably and unfavorably? Or is there some other reason?
If a trader is (a) extremely confident that his method will work deliver optimum - or at least very significant - profit, provided that he keeps to his rules, and (b) plays only with money that he can afford to lose, why should psychology be an issue?
On a personal level, the reason that I am likely to fail in the "mindset" part of the equation is because I believe that every method (i.e. especially my own) is fallible, no matter how extensively I've tested it. I've heard plenty of stories of folk whose method had delivered consistent profit for several months, only to lose it all when market conditions became unfavorable. That is a major reason why I believe that randomness plus costs will end up defeating all but the best methods, regardless of how well the method is implemented (i.e. MM plus psychology). (Having said that, though, I totally agree that poor MM and/or "emotional" trading will ravage an account much more quickly than an impotent method).
Another reason why I believe in "randomness plus costs" as being an underestimated enemy emanates from the discussions I've read about EAs. A black box robot will remove emotion - supposedly the #1 enemy - from the equation completely. Yet I read that most EAs fail because they are incapable of adapting to the prevailing market conditions. Therefore I conclude that adaptability - the ability to adjust the rules, perhaps even on the fly, when necessary - must be an important element in any trading method. But doesn't rule-bending amount to a violation of one's plan, a form of indiscipline? Or put another way, how does one qualitatively define which rules may be justifiably bent, and which ones may not? And, assuming that the backtesting was extensive, why weren't the "new" rules, or the adjustments, coded into the original EA? If re-coding an EA, to keep it "in synch" with current market "conditions", is an on-going necessity that must be performed indefinitely, doesn't that point to a very high degree of randomness in price movement?
I understand the concept of optimization, and over-optimization (or "curve fitting"). If a method's profitability is too heavily dependent on the "numbers" - the way that indicators are calibrated - then it lacks "robustness". So what exactly do we mean by "robust"? As an example, an MA crossover system is basically saying "price has been moving in the same direction long enough to cause the crossover, hence a 'trend' is apparently developing, therefore I assume that it's going to further continue long enough to (i) overcome costs and (ii) exceed the profit that's forfeited while waiting for the MAs to cross in the opposite direction". To whatever extent that trends are equally likely to end at any one point as any other, the premise is flawed, no matter how the MAs are calibrated.
What about trading pullbacks in trends?
OK, mathematically, if we make completely random entries and exits, our only enemy is costs, but (regardless of how randomly or not prices move) costs are enough to ensure eventual ruin. Now, as kiwi_trader points out, that when a trend exists, the impulse moves (that result in the creation of the higher highs in an uptrend; lower lows in a downtrend) are necessarily longer than the intervening retracements. Hence, for as long as the "trend" reliably exists, we improve our probabilities by trading in the direction of the trend. We know that trends involve profit-taking, and (also possibly for a variety of other reasons) there are retracements along the way, after which the trend resumes..... until eventually it doesn't
. These retracements result in an "oversold"/"overbought" condition occurring on a lower timeframe. So by entering during the pullback, we are entering at a "better" price (which, everything else being equal, will increase profit should the trade be a winner, and decrease loss, if a loser, i.e. the trend breaks down). Then we delay our actual entry until we see evidence (e.g. candle pattern; or movement on lower timeframe) that the "trend" is re-asserting itself, i.e. the retracement is over. So we have theoretically improved probability (beyond random entry/exit) on three counts:1. trading in the direction of the trend - but only to whatever extent a "trend" (e.g. higher highs and lows) is likely to continue;
2. by entering when a lower timeframe (wave) is OB/OS, which also improves potential RR - but only to what extent the OB/OS-ness is accurately picking a potential reversal point (end of retracement);
3. by delaying entry until the trend resumes - supposedly adds validity that the retracement has completed.
A long-winded ramble (sorry!), but the question is: Given that this method revolves around a "general principle", rather than "correctly" calibrated indicators, does this in itself make it any more "robust"? Or is the way in which we are quantifying the trend in point 1, and measuring the OB/OS-ness in point 2, ultimately amount to a form of calibration?
We could likewise attempt to dissect the rationale behind breakout trades, or fading extremes, and so on, but I think that's more than enough rambling from me for now. Good luck to anyone who actually reads all of this!!
David
- #66
- Oct 21, 2007 3:28pm Oct 21, 2007 3:28pm
I wasn't sure what you meant in your weight lifting analogy or how that related to the subject of trends.
Sometimes some threads become so full of opinions that the discussion deviates into other parallel discussions. You should be proud lol.
Gooner, superb post that I'm sure that virtually all of us can relate to.
A Blackjack pro will always double down whenever he holds 11 versus the dealer's 6. Even if he lost the last ten times doing so, he will continue to make the mathematically correct play, because he can be certain that - no matter what has happened in the recent past - the odds are significantly in his favor upon playing that way.
So why is it so much more difficult to "keep to one's rules" when trading? Is it because, whereas the Blackjack pro knows (given 11 vs 6) that the odds are in his favor, that the trader - no matter how extensive his backtesting, or how exhaustive his sample size - knows that probabilities can't be precisely calculated, and is therefore constantly haunted by the idea that his test results may be statistically uncertain? Or is it because, whereas a Blackjack hand is completed almost instantly, leaving little time to panic or agonize, the trader becomes spooked watching price dance up and down, favorably and unfavorably? Or is there some other reason?
If a trader is (a) extremely confident that his method will work deliver optimum - or at least very significant - profit, provided that he keeps to his rules, and (b) plays only with money that he can afford to lose, why should psychology be an issue?
On a personal level, the reason that I am likely to fail in the "mindset" part of the equation is because I believe that every method (i.e. especially my own) is fallible, no matter how extensively I've tested it. I've heard plenty of stories of folk whose method had delivered consistent profit for several months, only to lose it all when market conditions became unfavorable. That is a major reason why I believe that randomness plus costs will end up defeating all but the best methods, regardless of how well the method is implemented (i.e. MM plus psychology). (Having said that, though, I totally agree that poor MM and/or "emotional" trading will ravage an account much more quickly than an impotent method).
Another reason why I believe in "randomness plus costs" as being an underestimated enemy emanates from the discussions I've read about EAs. A black box robot will remove emotion - supposedly the #1 enemy - from the equation completely. Yet I read that most EAs fail because they are incapable of adapting to the prevailing market conditions. Therefore I conclude that adaptability - the ability to adjust the rules, perhaps even on the fly, when necessary - must be an important element in any trading method. But doesn't rule-bending amount to a violation of one's plan, a form of indiscipline? Or put another way, how does one qualitatively define which rules may be justifiably bent, and which ones may not? And, assuming that the backtesting was extensive, why weren't the "new" rules, or the adjustments, coded into the original EA? If re-coding an EA, to keep it "in synch" with current market "conditions", is an on-going necessity that must be performed indefinitely, doesn't that point to a very high degree of randomness in price movement?
I understand the concept of optimization, and over-optimization (or "curve fitting"). If a method's profitability is too heavily dependent on the "numbers" - the way that indicators are calibrated - then it lacks "robustness". So what exactly do we mean by "robust"? As an example, an MA crossover system is basically saying "price has been moving in the same direction long enough to cause the crossover, hence a 'trend' is apparently developing, therefore I assume that it's going to further continue long enough to (i) overcome costs and (ii) exceed the profit that's forfeited while waiting for the MAs to cross in the opposite direction". To whatever extent that trends are equally likely to end at any one point as any other, the premise is flawed, no matter how the MAs are calibrated.
What about trading pullbacks in trends?
OK, mathematically, if we make completely random entries and exits, our only enemy is costs, but (regardless of how randomly or not prices move) costs are enough to ensure eventual ruin. Now, as kiwi_trader points out, that when a trend exists, the impulse moves (that result in the creation of the higher highs in an uptrend; lower lows in a downtrend) are necessarily longer than the intervening retracements. Hence, for as long as the "trend" reliably exists, we improve our probabilities by trading in the direction of the trend. We know that trends involve profit-taking, and (also possibly for a variety of other reasons) there are retracements along the way, after which the trend resumes..... until eventually it doesn't
1. trading in the direction of the trend - but only to whatever extent a "trend" (e.g. higher highs and lows) is likely to continue;
2. by entering when a lower timeframe (wave) is OB/OS, which also improves potential RR - but only to what extent the OB/OS-ness is accurately picking a potential reversal point (end of retracement);
3. by delaying entry until the trend resumes - supposedly adds validity that the retracement has completed.
A long-winded ramble (sorry!), but the question is: Given that this method revolves around a "general principle", rather than "correctly" calibrated indicators, does this in itself make it any more "robust"? Or is the way in which we are quantifying the trend in point 1, and measuring the OB/OS-ness in point 2, ultimately amount to a form of calibration?
We could likewise attempt to dissect the rationale behind breakout trades, or fading extremes, and so on, but I think that's more than enough rambling from me for now. Good luck to anyone who actually reads all of this!!
David
I don't know about you, but I am always at odds with myself about many aspects of personal psychological behaviour. Whether it's a simple matter of refusing that extra downfall drink or not being able to bite your tongue when an obscure relative or (in your mind) interloping acquaintance presses the wrong "pet hate" button. I am constantly baffled at my inability to stand back, take a deep breath and move on...
I am in awe of traders who can discipline themselves sufficiently well to see their strategies succeed through the good and bad trades. I think I am getting there, but I also know that I am potentially one poor trade away from oblivion...
My current strategy (which seems to be working) is to take out profit each week that I manage to make some as long as I leave the bank in a healthier state than the previous balance (post previous deduction). The theory being to gradually compound the bank whilst seeing a return for my efforts. How long my spreadbetting company "Finspreads" will allow me to do this, assuming I continue to be successful, is anyones guess..
I also try desperately hard not to be in trades that are going wrong near to major news announcements. Exiting these when you are in a losing situation is difficult, yet vital a) to stay sane and b) to preserve your trading capital.
Anyway, I hope some day to be able to understand all the points you and your mathematically minded posters make.
Good Luck...
Gooner
- #69
- Oct 21, 2007 4:52pm Oct 21, 2007 4:52pm
- Joined Jul 2006 | Status: Charts + PA > * | 3,251 Posts
Dear lord.
Saw what happened to this thread..(really great chart + approach btw benssol, shows the difference)...
Okay, I'm gonna ruin myself by saying this(i.e. basically offend eveyone, likely mostly those that prefer to discuss what to do instead of doing it), but it's true and here goes:
You should all just stop posting, use the same energy and effort and focus it into chart time and live (or real time demo) trading and just wait another 6 months, then rethink whatever concepts you thought you had.
If nothing changed and you got nowhere, give it another 6, then if nothing changed - quit. Well, heck, at least it's an approach! One scenario at least.
I 've realized forum time(outside of VERY, VERY specific threads - and even those will be superceded eventually) can be so much more detrimental to trading than anything..the influence is just bad, the energy that's diverted...I just don't know..
Every minute put into posting would very likely lead to extensively bigger benefit if put into trading practice instead as long as one isn't "there" yet.(Then, after "getting there" you get incredibly bored waiting for the market to finally do what you entered it for hours ago and go on the forum prowl again perhaps..)
Dear lord...from trends don't exist, exist only in hindsight, exist in 50 forms and all timeframes over the random walk and Heisenberg back to Newton...yea there we go
I wouldn't feel one inch a better trader after one post nor after 50 posts or threads like this. I used to think that, probably. I think everyone has that head phase. Gotten over that, I hope.
However spend a week glued to the charts, alone and review what you did and why and you will get something out of it, go somewhere...so, on topic:
"If you thought that talking about forex get's you anywhere, you're wrong." You need to do it.
I guess this whole post will most likely sound insanely offensive(telling people what to do! my! tsk! the cheek! etc) and it's okay that it does.
It will do so because I believe it's true for a very large part and that always is uncomfortable, because knowingly or not, most people don't want to trade, they want to talk about (their)trading. Perhaps it's a social phenomenon? (maybe worth a sociological study..heheh..)
Btw, hanover(have to address you, your concerns just stand out so bad) - it's not a question of science. If you want to believe the market's are evil, negative sum games, hard to master, random or whichever you choose, you can find proof for whatever you wan't or against it, it doesn't matter.
There is only one thing that does: your personal belief and conviction. Because that's the only thing you will ever be able to trade, no matter what you define as real or not on paper and in 5000 attempts at a thesis. You will only be able to be and do yourself - and that's really all there is and the whole secret.
I will repeat what I found most important and then stop wasting everyone's time and this poor thread's space:
As a trader, one can't change the market - but one can change oneself.
That's really all you need to realize(as in in your head and then in turn in your actions) to get from being a losing trader to a winning one.
Best to everyone, please don't take offense, it's just a sort of outburst against all this "what the heck is everyone always trying to outtalk the market when they could be outtrading it instead...what ARE we here for?" feeling I'm getting lately when skimming over threads...really, don't take it as harsh as it sounds, but as serious a thought provocation as it should be...
Take care and a good TRADING week to everyone,
SeekingLight
Saw what happened to this thread..(really great chart + approach btw benssol, shows the difference)...
Okay, I'm gonna ruin myself by saying this(i.e. basically offend eveyone, likely mostly those that prefer to discuss what to do instead of doing it), but it's true and here goes:
You should all just stop posting, use the same energy and effort and focus it into chart time and live (or real time demo) trading and just wait another 6 months, then rethink whatever concepts you thought you had.
If nothing changed and you got nowhere, give it another 6, then if nothing changed - quit. Well, heck, at least it's an approach! One scenario at least.
I 've realized forum time(outside of VERY, VERY specific threads - and even those will be superceded eventually) can be so much more detrimental to trading than anything..the influence is just bad, the energy that's diverted...I just don't know..
Every minute put into posting would very likely lead to extensively bigger benefit if put into trading practice instead as long as one isn't "there" yet.(Then, after "getting there" you get incredibly bored waiting for the market to finally do what you entered it for hours ago and go on the forum prowl again perhaps..)
Dear lord...from trends don't exist, exist only in hindsight, exist in 50 forms and all timeframes over the random walk and Heisenberg back to Newton...yea there we go
I wouldn't feel one inch a better trader after one post nor after 50 posts or threads like this. I used to think that, probably. I think everyone has that head phase. Gotten over that, I hope.
However spend a week glued to the charts, alone and review what you did and why and you will get something out of it, go somewhere...so, on topic:
"If you thought that talking about forex get's you anywhere, you're wrong." You need to do it.
I guess this whole post will most likely sound insanely offensive(telling people what to do! my! tsk! the cheek! etc) and it's okay that it does.
It will do so because I believe it's true for a very large part and that always is uncomfortable, because knowingly or not, most people don't want to trade, they want to talk about (their)trading. Perhaps it's a social phenomenon? (maybe worth a sociological study..heheh..)
Btw, hanover(have to address you, your concerns just stand out so bad) - it's not a question of science. If you want to believe the market's are evil, negative sum games, hard to master, random or whichever you choose, you can find proof for whatever you wan't or against it, it doesn't matter.
There is only one thing that does: your personal belief and conviction. Because that's the only thing you will ever be able to trade, no matter what you define as real or not on paper and in 5000 attempts at a thesis. You will only be able to be and do yourself - and that's really all there is and the whole secret.
I will repeat what I found most important and then stop wasting everyone's time and this poor thread's space:
As a trader, one can't change the market - but one can change oneself.
That's really all you need to realize(as in in your head and then in turn in your actions) to get from being a losing trader to a winning one.
Best to everyone, please don't take offense, it's just a sort of outburst against all this "what the heck is everyone always trying to outtalk the market when they could be outtrading it instead...what ARE we here for?" feeling I'm getting lately when skimming over threads...really, don't take it as harsh as it sounds, but as serious a thought provocation as it should be...
Take care and a good TRADING week to everyone,
SeekingLight
Trust price. Know yourself.
- #70
- Oct 21, 2007 6:16pm Oct 21, 2007 6:16pm
- | Joined Apr 2005 | Status: Cruising The Outer Rim... | 2,175 Posts
- #72
- Oct 21, 2007 6:53pm Oct 21, 2007 6:53pm
- Joined Sep 2006 | Status: ... | 8,112 Posts
..... You should all just stop posting, use the same energy and effort and focus it into chart time.....
I’m not adamant about any of my views; they have evolved over time, and I hope that they will continue to grow and change, as I keep on learning. For me, a large part of the learning process involves the interchange of ideas. I have learned a lot from reading the posts on this forum.
Trading can be a lonely (and when things are going wrong, extremely frustrating) business, and I can’t see how the innate desire to express oneself is in itself wrong. There is plenty of downtime between potential trading setups, so I don’t believe that "trading time and energy" is being wasted in posting one’s opinions, whatever their merit.
There are many ways to approach trading. Schwager’s Market Wizard books include success stories of some who have used scientific, statistical or probability-based approaches.
Personally, I have found attaining consistent profit extremely difficult. Apparently I’m not alone, as the vast majority of newbies, like myself, lose. Of course I may be wrong – perhaps I’m missing something obvious – but I don’t believe that my problems exist in my head, or that if I somehow "changed myself" that my results would improve. I find it easy to make disciplined entries and exits, it’s just that I find myself getting their timing wrong. When I enter long, the market tends to fall, more often than not; when I enter short, it rises. To me, that’s a methodological, rather than a "mindset" related, issue. I can’t see how prices are going to move differently, just because I choose to think more positively.
If you are one of the minority who are making profit consistently, then I congratulate and admire you, and I wish you every continued success (as long as it's not my money that you're taking!!
)David
Trend being or not being your friend is not a matter of rhetorics, it's a matter of facts and figures. Make a very simple test.
Draw, say, 24 period moving averages on hourly, daily, weekly, and monthly GBPUSD charts.
Place a position every day at London open with, say, 80 p. SL and 130 p. TP.
If your position is opened in a random direction, your expected daily gain is -9 p.
If your position is opened in the direction of the hourly trend (the hourly MA pointing North), your expected daily gain is +7 p.
If your position is opened in the direction of the hourly AND the daily trend, your expected daily gain is +10 p.
If your position is opened in the direction of the hourly AND the daily AND the weekly trend, your expected daily gain is +13 p.
If your position is opened in the direction of the hourly AND the daily AND the weekly AND the monthly trend, your expected daily gain is +18 p.
Wow, we have a fantastic trading system here! I'll call it Ilan's Quadruple (IQ) method, publish it in the trading systems section and attract thousands of admirers! Or, better still, I'll not publish it. If everyone will use it, it'll stop working. :-)
Now on a serious note. One can trade with the trend, against the trend, regardless of the trend and, probably, be profitable. However, all other things being equal, a trend is your friend. Like many other truisms, this one is true.
Looks like the sleeping bears woke and toke over the trend thingies.
I apologize if I seem like I'm making things complicated.
Trends are not random walks, though they can give that appearance. There is a confluence of factors that go into a trend. These factors are introduced by every player in the market. Some players influence these factors more than others.
When all these factors are in balance, you have a fairly stable state in the market. When these factors go out of balance, a trend forms.
In fact, every single one of us is a variable in the market, which also are comprised by yet more variables (risk levels, greed, sensitivity, etc.). Some variables play a bigger part than others. Some variables are influenced by certain events more than others. It's a fascinating and complex system to watch and be a part of.
Even the classic coin toss example is not truly random in real life. Variables such as thumb flexion, air resistance, and gravity all play a part in determining whether or not heads or tails will show up. It may appear random but with enough practice one could indeed bias the results.
Of course, our problem is trying to determine when and how long a trend will form. Unfortunately, this would require a system capable of calculating billions of variables with intimate knowledge of all the traders in the system.
The parallels to physics is that a perfect model of say, universal expansion, would also require a similar amount of knowledge about the structures in the universe.
Since perfect knowledge of either system is impossible, both physicists and traders attempt to use a less perfect model. Instead of tracking everything (impossible) we make educated guesses about what we don't know and attempt to analyze key parameters that have the greatest importance.
Examples of the key things we utilize to help us quantify the market are price, volume, an news impacts. We use these variables under the assumption that they reflect a majority of the market sentiment. In doing so, we attempt to make certain predictions about the market. This is similar to a physicist trying to predict the flow of a liquid, where trillions of atoms are interacting with one another.
Unfortunately, due to the huge number of influences both large and small there is no way to be right 100% of the time. No matter how fine you're market measurements may be, there will always be some amount of uncertainty. There is always a chance the the right amount of variables will coincide at the wrong time and change your clear trend to a flatline. This the same principle behind quantum mechanics. You may be able to know some things about a particle, but it is impossible to know everything about a particle.
In an effort to utilize this line of reasoning, I created a simple indicator based on closing price deltas as a form of momentum under the assumption that closing price accurately reflects the current forcings in the market. This is just a start, but the indicator is fairly decent.
1. Take N closing price deltas: (x0 - x1), (x1-x2), etc.
2. Take an EMA of the price deltas
3. Take an EMA of the EMA.
4. Plot as a histogram, with increasing lines colored green and decreasing lines colored red.
This results is a histogram that pretty accurately marks up and down turns. It's still in the testing phase, but values between 21 and 35 periods seem to work the best. If you get a couple of the same colored bars in a row it's a very good sign that a trend change is about to happen.
In comparison to the Awesome Oscillator, this one seems to respond slightly better to changes in momentum. This one can also have longer MA periods. The longer the MA period, the longer the trend it detects. The shorter the period, the shorter the trend. But make the period too short and you'll end up just getting the noise.
The MA should hypothetically be varied based on volatility (like ATR) in order to filter out noisy fluctuations, but I'm still working on the relationship. I've also been thinking about adding volume in as well as a possible impulse factor but like I said I'm starting simply.
Mini-Me, I haven't posted enough yet to be able to respond to your message but I will once I can.
I've got to stop with these long posts.
~X~
Trends are not random walks, though they can give that appearance. There is a confluence of factors that go into a trend. These factors are introduced by every player in the market. Some players influence these factors more than others.
When all these factors are in balance, you have a fairly stable state in the market. When these factors go out of balance, a trend forms.
In fact, every single one of us is a variable in the market, which also are comprised by yet more variables (risk levels, greed, sensitivity, etc.). Some variables play a bigger part than others. Some variables are influenced by certain events more than others. It's a fascinating and complex system to watch and be a part of.
Even the classic coin toss example is not truly random in real life. Variables such as thumb flexion, air resistance, and gravity all play a part in determining whether or not heads or tails will show up. It may appear random but with enough practice one could indeed bias the results.
Of course, our problem is trying to determine when and how long a trend will form. Unfortunately, this would require a system capable of calculating billions of variables with intimate knowledge of all the traders in the system.
The parallels to physics is that a perfect model of say, universal expansion, would also require a similar amount of knowledge about the structures in the universe.
Since perfect knowledge of either system is impossible, both physicists and traders attempt to use a less perfect model. Instead of tracking everything (impossible) we make educated guesses about what we don't know and attempt to analyze key parameters that have the greatest importance.
Examples of the key things we utilize to help us quantify the market are price, volume, an news impacts. We use these variables under the assumption that they reflect a majority of the market sentiment. In doing so, we attempt to make certain predictions about the market. This is similar to a physicist trying to predict the flow of a liquid, where trillions of atoms are interacting with one another.
Unfortunately, due to the huge number of influences both large and small there is no way to be right 100% of the time. No matter how fine you're market measurements may be, there will always be some amount of uncertainty. There is always a chance the the right amount of variables will coincide at the wrong time and change your clear trend to a flatline. This the same principle behind quantum mechanics. You may be able to know some things about a particle, but it is impossible to know everything about a particle.
In an effort to utilize this line of reasoning, I created a simple indicator based on closing price deltas as a form of momentum under the assumption that closing price accurately reflects the current forcings in the market. This is just a start, but the indicator is fairly decent.
1. Take N closing price deltas: (x0 - x1), (x1-x2), etc.
2. Take an EMA of the price deltas
3. Take an EMA of the EMA.
4. Plot as a histogram, with increasing lines colored green and decreasing lines colored red.
This results is a histogram that pretty accurately marks up and down turns. It's still in the testing phase, but values between 21 and 35 periods seem to work the best. If you get a couple of the same colored bars in a row it's a very good sign that a trend change is about to happen.
In comparison to the Awesome Oscillator, this one seems to respond slightly better to changes in momentum. This one can also have longer MA periods. The longer the MA period, the longer the trend it detects. The shorter the period, the shorter the trend. But make the period too short and you'll end up just getting the noise.
The MA should hypothetically be varied based on volatility (like ATR) in order to filter out noisy fluctuations, but I'm still working on the relationship. I've also been thinking about adding volume in as well as a possible impulse factor but like I said I'm starting simply.
Mini-Me, I haven't posted enough yet to be able to respond to your message but I will once I can.
I've got to stop with these long posts.
~X~
- #80
- Oct 22, 2007 11:20am Oct 22, 2007 11:20am
- | Joined May 2007 | Status: Seasoned Trader | 820 Posts
Trends are not random walks, though they can give that appearance.
My question here is Does absolute ramdomness ever occur in the physical world?
I've got to stop with these long posts.
???
~X~
What we label the variables matters not. It is therefore more important to understand the variables and their possible effects upon the market.
Note: The trend is not my friend! The only friend I desire in trading is the market. It is best to become one with the market and nurture a long lasting relationship. Skrew the trend!