Inescapable Truths | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trading Truths 79 replies

Popular Trading Tips: Truths or Myths 36 replies

A collection of truths and proto-truths 5 replies

Attachments: Inescapable Truths
Exit Attachments

Inescapable Truths

  • #101
  • Nov 18, 2009 8:02pm Nov 18, 2009 8:02pm
Quoting hanover
So which is true:
that price is equally likely to rise X pips as it is to fall X pips
OR
that price is equally likely to rise X% as it is to fall X%
Neither are true. The assumption is flawed.
  • #102
  • Nov 19, 2009 4:34am Nov 19, 2009 4:34am
Quoting hanover
Many thanks, I found the link back in post #57.

I suspect that the concept is a cleverly constructed illusion, whose basis is that price is equally as likely to rise 1% as it is to fall 1% (a geometric progression).

If we base the upward/downward steps on an arithmetic progression, i.e. a rise or fall of 0.01, and then sum the final column, the overall expectancy is zero (which is what I would expect).

So which is true:
that price is equally likely to rise X pips as it is to fall X pips
OR
that price is equally likely to rise X% as...
Yes, if you read the paragraph under "Constructing a random walk," they tell you that after each period (each day in this case), price moves up 1% or down 1% randomly. There's a 50% chance it moves up, 50% chance it moves down. They based it on a coin flip. So that's not an illusion, that's exactly what a classic random walk is: 1. Start at the current position. 2. Go up or down randomly. The outcome has nothing to do with what happened in the past. 3. Repeat.

Now, about your question: Is price moving a fixed distance or a fixed percentage change? Well in their example, they explain that it's moving a fixed percentage change.

If you don't like that idea, you can assume that price is moving a fixed distance. Then you just compound your account every day, i.e. change the position size so that when price moves that fixed distance, it will amount to either a 1% increase or 1% decrease in your account. This way, it will match the example again.

Of course in real life, the market doesn't move with a fixed distance or fixed percent change every day. Also, it would seem that prices are not completely independent of each other, i.e. today's price is affected by past prices.

Hmm...but I wonder, if we used a fixed TP and fixed SL, and used the same trading rules... ...Doh! Damn transaction costs!
  • #103
  • Nov 19, 2009 9:43am Nov 19, 2009 9:43am
Quoting mathematician
Yes, if you read the paragraph under "Constructing a random walk," they tell you that after each period (each day in this case), price moves up 1% or down 1% randomly. There's a 50% chance it moves up, 50% chance it moves down. They based it on a coin flip. So that's not an illusion, that's exactly what a classic random walk is: 1. Start at the current position. 2. Go up or down randomly. The outcome has nothing to do with what happened in the past. 3. Repeat.

The problem is that they cheated. When price goes up, they multiply it by 1.01, so add 1%, however when price goes down, they divide by 1.01. However this is not the same as a 1% downward move - it's a 0.99% downward move. So in essence they have a biased random walk, with a larger upward moves. If this is done correctly, i.e. when you have a downward move, you multiply by 0.99 (rather than dividing by 1.01), you will get 0 with their strategy.

It doesn't matter whether the series is geometric or arithmetic, what matters is whether it is biased or not.
  • #104
  • Nov 19, 2009 9:57am Nov 19, 2009 9:57am
bars are fun. In several threads Renko bar trading is discussed. I direct your attention to the info about probability of continuation. UDU means Up down up. In this progression, the work of several participants in one of the threads shows that in the case of E/J, using .10 Renko bars, the liklihood of U, a one bar continuation, rather than D, a two bar reversal, is >70%, 72%, if I remember correctly. The trial was > 400 observations. The closing price of series UDU or DUD, has, has to move 1/2 the distance for a continuation as it does for a reversal. In this case 10 pips vs 20pips. At this point I must ask...Why is the outcome, continuation > 66.6%.
Vincit qui se vincit.
  • #105
  • Nov 19, 2009 12:46pm Nov 19, 2009 12:46pm
Quoting mathematician
(snip)

For me it really emphasizes an important point: Good trading is not about predicting where prices are going. It's about taking a position and managing it.

Yes, I completely agree here. Elements of subjectivity and judgment are required for determining not only what direction to favor, but also where to enter and exit, and how to protect profits during the open position. It's a game of approximates. If no subjective judgement is used, one could end up like those poor people over in the programming discussion board - attempting to use an objective re-interpretation (indicator output) of something inherently unpredictable to make subjective choices without actually making the subjective choices.
  • #106
  • Nov 19, 2009 12:49pm Nov 19, 2009 12:49pm
Quoting Lionstrike
Yes, I completely agree here. Elements of subjectivity and judgment are required for determining not only what direction to favor, but also where to enter and exit, and how to protect profits during the open position. It's a game of approximates. If no subjective judgement is used, one could end up like those poor people over in the programming discussion board - attempting to use an objective re-interpretation (indicator output) of something inherently unpredictable to make subjective choices without actually making the subjective choices.
Completely wrong, there is nothing subjective about an edge, anything else is gambling.
The breaking of a wave cannot explain the whole sea.
  • #107
  • Nov 19, 2009 12:51pm Nov 19, 2009 12:51pm
Here's a truth that I've discovered from experience.

No amount of hedging or scaling in and out of positions will give you any kind of inherent edge. You can't circumvent the need for good timing with these methods. At the end of the day you need to be able to take a reasonable loss and to have a good sense of when to take a profit and when to simply trail.
  • #108
  • Edited 1:04pm Nov 19, 2009 12:53pm | Edited 1:04pm
Quoting Craig
Completely wrong, there is nothing subjective about an edge, anything else is gambling.
How exactly are you quantifying an "edge"? In the long run, given enough data and enough sample tries through the data, the only edge you can give yourself is through subjective judgement.

If there was any inherent edge to be had from OBJECTIVE re-interpretations of price (indicator outputs) - which I might add can be helpful as reference points - then one could simply program a computer to take certain indicator outputs as entries and certain others as exits. But that isn't possible - there's no subjectivity in that. The indicator outputs in a consistent, objective way, applying its mathematical calculation the same way each time. You're not taking any risks when you rely on indicators alone, without any subjective judgment, and the market rewards intelligent risk taking.
  • #109
  • Nov 19, 2009 1:06pm Nov 19, 2009 1:06pm
Quoting Lionstrike
How exactly are you quantifying an "edge"?
Historical expectation & real-time t-test.

Quoting Lionstrike
In the long run, given enough data and enough sample tries through the data, the only edge you can give yourself is through subjective judgement.
Wrong, data is for testing hypothesis, not for randomly trying things.
Something works or it does not, there is no 'works if I use my special judgment'.
The breaking of a wave cannot explain the whole sea.
  • #110
  • Edited 1:38pm Nov 19, 2009 1:14pm | Edited 1:38pm
Quoting Craig
Historical expectation & real-time t-test.


Wrong, data is for testing hypothesis, not for randomly trying things.
Something works or it does not, there is no 'works if I use my special judgment'.

Ok, then I challenge you to provide proof, over 1,000 runs in real-time of something that works without subjective judgement involved.

Or at least provide proof of something that gives you a "historical expectation". How can price give you some expectation of itself? That's BS. If there was any inherent expectation, then one could simple take moving average crossovers, for example, all the time.

No objective reinterpretation of price itself can give you any historical expectation about price in the long run. Can I drive a car by saying "Ok, I'll look in the rearview mirror and if I don't see any cars there or to my left or right then I can turn the steering wheel 2 inches and step on the gas half an inch?" Of course not, that's pure bs.

....
My point, Craig, is this - Haven't you ever done something simple over and over again until you get an instinctual feel for it? Something mundane - maybe play a sports video game over and over until pressing the buttons becomes second-nature and your timing is impeccable? Or practice something like juggling enough hours to the point where it's easy? Or something like martial arts moves? This is a skill acquired through the same application of practice. Looking at the same screens and same lines and dots for a certain amount of hours will allow one to develop an instinctual feel for it.

That's my opinion. You certainly are free to hunt for some hidden objective mathematical edge though and I'm sure your broker appreciates it. I think it's fool's gold. But good luck, nonetheless.
  • #111
  • Nov 19, 2009 1:48pm Nov 19, 2009 1:48pm
Quoting Lionstrike
Ok, then I challenge you to provide proof, over 1,000 runs in real-time of something that works without subjective judgement involved.
Nice try, but you surly must be joking, I'm not giving away years of work.

Quoting Lionstrike
Or at least provide proof of something that gives you a "historical expectation". How can price give you some expectation of itself? That's BS. If there was any inherent expectation, then one could simple take moving average crossovers, for example, all the time.
I have no idea what you mean here, perhaps I was vague, by "historical expectation" I mean, the results of my testing an effect on data yielded a positive expectation, nothing too complex there I hope.

Quoting Lionstrike
No objective reinterpretation of price itself can give you any historical expectation about price in the long run. Can I drive a car by saying "Ok, I'll look in the rearview mirror and if I don't see any cars there or to my left or right then I can turn the steering wheel 2 inches and step on the gas half an inch?" Of course not, that's pure bs.
That's true for data mining, I don't data mine. Statistical biases persist.

Quoting Lionstrike
My point, Craig, is this - Haven't you ever done something simple over and over again until you get an instinctual feel for it? Something mundane - maybe play a sports video game over and over until pressing the buttons becomes second-nature and your timing is impeccable? Or practice something like juggling enough hours to the point where it's easy? Or something like martial arts moves? This is a skill acquired through the same application of practice. Looking at the same screens and same lines and dots for a certain amount of hours will allow one...
Err...of course I have, I just don't think it applies to trading, at least not in the way you seem to imply. Anyway, we are just going to have to beg to differ.
The breaking of a wave cannot explain the whole sea.
  • #112
  • Nov 19, 2009 1:58pm Nov 19, 2009 1:58pm
Craig, if you were referring to trends and momentum, I do think momentum can and DOES provide an edge (there clearly exists price dependency in the market)...what I was saying is that I don't think the edge that it provides can be calculated reliably mathematically over time, outside of the concept of subjectivity.

There's no doubt that both momentum and unpredictability can both help and hurt one's trading. So in trading momentum can be either long or short. So what I do, and I'm sure you do something different, is simply and subjectively decide on a direction to favor and take trades in the direction only, until I feel the trend has changed and then I take trades in the other direction only, etc. So the momentum provides the edge, yes, but I subjectively make that choice about when the trend changes. The rest is simply in-trade management and dealing with unpredictability.

Does this not make sense?
  • #113
  • Edited 4:04pm Nov 19, 2009 3:46pm | Edited 4:04pm
I find your posting somewhat muddled, but you seem to be mixing up a lot of different ideas. Let us strip the argument back somewhat, I have attached a excel graph of close prices which I have normalized so as not to give away where they came from. I want you to apply your 'subjective feel for uncertainty' or whatever you call it to this chart and tell me what you think will happen next.
Edit: Anybody else feel free to chime in with their own analysis.
Attached Image (click to enlarge)
Click to Enlarge Name: chart.jpg Size: 67 KB
The breaking of a wave cannot explain the whole sea.
  • #114
  • Edited 5:07pm Nov 19, 2009 4:55pm | Edited 5:07pm
Quoting Craig
I find your posting somewhat muddled, but you seem to be mixing up a lot of different ideas. Let us strip the argument back somewhat, I have attached a excel graph of close prices which I have normalized so as not to give away where they came from. I want you to apply your 'subjective feel for uncertainty' or whatever you call it to this chart and tell me what you think will happen next.
Edit: Anybody else feel free to chime in with their own analysis.
You'd like me to take a wild guess at direction based upon some cherry-picked example of close prices? You must think me a fool!

Anyway, I don't trade like that. I don't really care what will happen next. I don't make my trading decisions based upon what I think will happen, only by what is happening relative to what happened. Good luck trading though, I think we're done here.
  • #115
  • Nov 19, 2009 5:10pm Nov 19, 2009 5:10pm
Quoting Craig
Edit: Anybody else feel free to chime in with their own analysis.
My 'subjective feel for uncertainty' tells me that it's a chart of a random walk.
  • #116
  • Nov 19, 2009 5:11pm Nov 19, 2009 5:11pm
Quoting Lionstrike
You'd like me to take a wild guess at direction based upon some cherry-picked example of close prices? You must think me a fool!
LOL, Oh, come on, look at that trend! What more do you want?

Quoting Lionstrike
Anyway, I don't trade like that. I don't really care what will happen next. I don't make my trading decisions based upon what I think will happen, only by what is happening relative to what happened.
Can't argue with that logic
The breaking of a wave cannot explain the whole sea.
  • #117
  • Nov 19, 2009 5:14pm Nov 19, 2009 5:14pm
Quoting euclid
My 'subjective feel for uncertainty' tells me that it's a chart of a random walk.
The prize goes to euclid!
Now is it a random walk I generated, or one I pulled out of a tick database?
If so, why?
The breaking of a wave cannot explain the whole sea.
  • #118
  • Nov 19, 2009 6:45pm Nov 19, 2009 6:45pm
Quoting Craig
LOL, Oh, come on, look at that trend! What more do you want?



Can't argue with that logic
Hmmm, now you have my curiosity piqued. Earlier you wrote:

"Completely wrong, there is nothing subjective about an edge, anything else is gambling."

Then you go ahead and put up a chart of a random walk. Just what point are you trying to prove, if you are trying to prove a point about price action and not simply trying to bait me into some pathetic debate point?

How exactly would you gain an edge, sir, over a random walk? Or , if you are not making that claim, what was the point of putting up a chart of a random walk of prices?
  • #119
  • Nov 19, 2009 6:53pm Nov 19, 2009 6:53pm
Quoting Craig
The prize goes to euclid!
Now is it a random walk I generated, or one I pulled out of a tick database?
If so, why?
I'll let you know my answer when I have received my prize.
  • #120
  • Nov 19, 2009 7:01pm Nov 19, 2009 7:01pm
Quoting Lionstrike
Hmmm, now you have my curiosity piqued. Earlier you wrote:

"Completely wrong, there is nothing subjective about an edge, anything else is gambling."

Then you go ahead and put up a chart of a random walk. Just what point are you trying to prove, if you are trying to prove a point about price action and not simply trying to bait me into some pathetic debate point?

How exactly would you gain an edge, sir, over a random walk? Or , if you are not making that claim, what was the point of putting up a chart of a random walk of prices?
I wish you would stop getting all overheated, it's not meant to be personal, but anyway the point is, if you (or anybody else who has been making assertions on this thread for that matter) can't tell me what is the difference between a random walk and a chart in any objective manner then how on earth can we have a whole thread of 'Inescapable Truths' given that we can't even define what we are talking about? As for edges in random processes, well...I'm not going to give everything away, check out the concept of 'pairs trading', not the bullshit correlation pairs trading but the proper co-integrated version, this should alert you to the fact that all noise is created equal.
The breaking of a wave cannot explain the whole sea.
0 traders viewing now