why have you accepted the price-predictability assumption? | Page 5 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

So I accepted a Prop Trading job in South Beach Miami 44 replies

Why you shouldn't show how much you have made in FX 35 replies

Incorporating Slippage Assumption in backtesting 4 replies

Market Predictability - Technical Signals 25 replies

Contesting the Accepted View of Bernanke 0 replies

Attachments: why have you accepted the price-predictability assumption?
Exit Attachments

why have you accepted the price-predictability assumption?

  • #81
  • Dec 15, 2013 10:57am Dec 15, 2013 10:57am
Quoting Turveyd
{quote} The random generated charts which look close to real, where momentum seems to play parts is due to the pseudo randomness of computers, which can't generate truely random numbers. Well the C64's could, they used there sound SID Chip and actually had REAL random numbers, everyone else had to use a seed and therefore Psuedo. See they've found a C64 like way using some other external variable, nice, used to be based on Time before. Set Dragon 32's Timer to 0 and same line load a game ( Timer stopped while loading ) and get the same security...
Well the most legit source of random numbers are radiation source or electric impulses in a chip.

But i found many articles on the net where you can generate random numbers on your own, for example if you have a microphone you can let it record your room default noise, and then use a decibel filter to split the sound file into 2 pieces and the mean of the volume will be the half point, and if the noise is louder then =1 if its softer then 0.

Or the CPU clock rate is somewhat random too.
______________

But to simulate a chart you must use true random numbers since pseudo random's will repeat after a given time, and we know that in the market there are no identical deja vu segments so, anyhow the market is constructed, if its random, it must be a true random sequence
"There's a sucker born every minute" - P.T. Barnum
  • #82
  • Dec 15, 2013 11:04am Dec 15, 2013 11:04am
Quoting Turveyd
{quote} SID chips reading was on the C64 truely random, some Noise thing on that site, would like more information, truely random seed = truely random number.
Ah I see. Sounds like more trouble than it's worth to me as pseudorandom numbers based on timestamps work quite well for almost all purposes, but taking a look at the wiki article on "truly" random numbers the way they do it makes enough sense.
  • #83
  • Dec 15, 2013 11:06am Dec 15, 2013 11:06am
Quoting BooG
@60minuteman (your post popped up as i was writing the replay) Yes it's 50/50.there's just 2 options up or down BUT, if you look at the picture attached to this reply you will see that CHANCES ARE the price will fall. And that's were you make money. So, the 50/50 turned into a, let's say, 75/25... market inefficiency at its very best To recap, markets are sometimes predictable. behind every trade there needs to be a reason. always keep it stupid simple (kiss). It's just up and down for god's sake stop making it black magic science. Have an...
Hi Boog, it is probably comforting to believe it's 75/25 and that it will fall, but the reality is that its 50/50 and it DID fall.

Everyone likes to build their strategies on a higher than 50% win rate, but the smart move is to make 50/50 work, then any chnce that it is 75/25 will only be a bonus.

best of luck dude!
tradewith60
  • #84
  • Dec 15, 2013 11:19am Dec 15, 2013 11:19am
Quoting BooG
Hi all! Very good topic idea/question. As simple as it sounds, there are some very technical and fancy math replies. (page 2-3). I study Engineering, (automation&compu science) and i was kinda lost in those distributions graphs. I don't think that they are useless, god forbid, I love graphs and quantifying anything that moves, but I do not think they give an edge to a trader. This being said, let's remember WHY the market MOVES. Well, the easy answer is: somebody buys/sells this particular asset.(i know there are more hidden reasons: cash flow,...
this is why I started this thread.....we all have assumptions, some of us very good assumptions some of us very bad.

HOW do you know that dojis work? do you have a statistic study on it? what is the probability in % terms that a decline will be followed by a doji? or a shooting star? or head and shoulders? what % of the time does a bull leg start after a bull flag?

same goes with all of our assumptions, if in fact they are NOT statistically significant then we are shooting darts in the dark aren't we?
AVT INVENIAM VIAM AVT FACIAM
  • #85
  • Dec 15, 2013 11:22am Dec 15, 2013 11:22am
Quoting 60minuteman
{quote} Everyone likes to build their strategies on a higher than 50% win rate, but the smart move is to make 50/50 work,
You are totally right and if you take a glance at my trader explorer, you will see my win rate is a newbish 55% (over 6 months), yet the profit stands at 700%. That means my trades start at 50/50 chances to succeed but, BECAUSE OF T.A. reasoning, the winners make a whole bunch more than the losers.
IMO this is the only way to make it in FX and achieve the all wanted "trading for a living" life-style.

Back to the topic, I once tried to write an EA that MUST work on a ECN broker. The idea behind it : set a random function that gives 1 and 0 (1=buy, 0=sell), make it open a trade at every new 1min candle. there is just one trade active at anytime. The trade will be closed after 30 seconds OR when the loss >0.1% account. As for the profit, i don't care. Unfortunatly i suck at coding mql4...but i would be so intrested as how the EA would perform...maybe some day.Or a very giving person writes it. just the backbone
  • #86
  • Dec 15, 2013 11:24am Dec 15, 2013 11:24am
Quoting the redlion
{quote} what is the probability in % terms that a decline will be followed by a doji? or a shooting star? or head and shoulders? what % of the time does a bull leg start after a bull flag?
Better than 50%. That coupled with good money management is enough.
It dosen't have to be dojis or patterns(i used dojis because they are easy to find on a chart and i'm lazy ). As long as it's SOMETHING that over a big data sample gives 51% right answers, it's a keeper.

You can change SOMETHING with : trendlines, SMA crossing, MACD divergences, a black&white dove that does a loop while flying. Whatever reason that you know from experience/watching it ,and that gives you higher that 50/50 chances.
  • #87
  • Dec 15, 2013 11:30am Dec 15, 2013 11:30am
Here is my proof of randomness with bias:

Attached Image (click to enlarge)
Click to Enlarge Name: eurcadh1.png Size: 21 KB


You can see 6 consecutive bullish candles, which have a probability of happening in a 50-50 market of: 1.5625%

But look 8 consecutive bullish bars in a short period of time , also if we suppose that this market had a 50-50 chance, this to happen would have a 0.390625 %

But both of them happening so close after one another must prove only 1 thing, (atleast) this segment of the market was not 50-50% in the first place, but it was an "unfair coin", biased upwards, also know as an uptrend.

And guess what, it was really an uptrend :

Attached Image (click to enlarge)
Click to Enlarge Name: uptrend.png Size: 27 KB


So in my point of view, an uptrend is nothing more than a biased, already random market, where the bias comes from fundamental factors.
A downtrend is the same but in the other direction, while a sideways movement is the perfect 50-50 situation where there is no bias/trend, so the market will take its real form of 50-50

What do you think guys?
"There's a sucker born every minute" - P.T. Barnum
  • #88
  • Dec 15, 2013 11:34am Dec 15, 2013 11:34am
Quoting BooG
{quote} Better than 50%. That coupled with good money management is enough.
No that's a belief, a irrational belief.

Okay H4 candles, move the start and end time around and you can make all the candles there are pretty much then when you understand the arbitrary nature of time then you'll understand all candles are randomly formed.

A candle which goes down say 30pips, then closes at the bottom then goes back up 30pips is exactly the same PA as a hanging man is it ?


It's the reading of many candles which gives you trend / direction not 1.
Nothing to it, but to do it!!! Stick to the plan FOOL!!!!
  • #89
  • Dec 15, 2013 11:43am Dec 15, 2013 11:43am
Quoting Turveyd
{quote} Okay H4 candles, move the start and end time around and you can make all the candles there are pretty much then when you understand the arbitrary nature of time then you'll understand all candles are randomly formed.
H4 candles are subject to the time your broker gives you. So unless both our brokers start at 00.00 GMT, your chart will be different from mine. This does not happen on H1. everyone, and I mean everyone sees the same candle formation.

But again, I used dojis in order to cover the concept of REASON behind the trade. the sell signal (in my case doji) could have been on "JackTheTrader"s chart a CCI bearish divergence...

REASON = a thing that based on historical data, gives you signals that are more accurate than a coin flip.
REASON = The trader's weapon that keeps that awful "randomness monster" away )
  • #90
  • Dec 15, 2013 11:51am Dec 15, 2013 11:51am
Quoting BooG
{quote} H4 candles are subject to the time your broker gives you. So unless both our brokers start at 00.00 GMT, your chart will be different from mine. This does not happen on H1. everyone, and I mean everyone sees the same candle formation. But again, I used dojis in order to cover the concept of REASON behind the trade. the sell signal (in my case doji) could have been on "JackTheTrader"s chart a CCI bearish divergence...

But even H1 it's not like the market / traders all over the world go right it's 13:00 lets draw this H1 candle stick pattern.

Candle Stick patterns have ZERO value move away quickly from using them.

Divergence LOL another myth, if you understood the math's as to what your seeing there, you'll also understand it's nothing, your basing trade decisions on randomness again.
Nothing to it, but to do it!!! Stick to the plan FOOL!!!!
  • #91
  • Dec 15, 2013 11:53am Dec 15, 2013 11:53am
Quoting Turveyd
{quote} But even H1 it's not like the market / traders all over the world go right it's 13:00 lets draw this H1 candle stick pattern. Candle Stick patterns have ZERO value move away quickly from using them. Divergence LOL another myth, if you understood the math's as to what your seeing there, you'll also understand it's nothing, your basing trade decisions on randomness again.
Independently, candles don't have a lot of value. In context, they can give you a good sign as to whether or not strong buying or selling is going on in a target zone though. It's more about the wicks and quick retreats than anything else.
  • #92
  • Dec 15, 2013 11:58am Dec 15, 2013 11:58am
Quoting BooG
{quote} You are totally right and if you take a glance at my trader explorer, you will see my win rate is a newbish 55% (over 6 months), yet the profit stands at 700%. That means my trades start at 50/50 chances to succeed but, BECAUSE OF T.A. reasoning, the winners make a whole bunch more than the losers. IMO this is the only way to make it in FX and achieve the all wanted "trading for a living" life-style.

your data is not newbish at all, it's exactly what I would expect from a trader that was trading in a profitable manner, win rates will go up and down but over time they will all be around 50%, wether you trade elliot wave, SR, bollock bands, whatever.... as soon as you accept this youre on the path.

You have obviously found A path, wether you did it by luck or craft is irrelevant... but well done anyway.
tradewith60
  • #93
  • Dec 15, 2013 11:58am Dec 15, 2013 11:58am
Quoting Turveyd
{quote} But even H1 it's not like the market / traders all over the world go right it's 13:00 lets draw this H1 candle stick pattern. Candle Stick patterns have ZERO value move away quickly from using them. Divergence LOL another myth, if you understood the math's as to what your seeing there, you'll also understand it's nothing, your basing trade decisions on randomness again.

Ok. So, why are we still here? It means there is just no way to trade the markets ,since they are all a random string of digits and we'll all end up with break even.

The thread starter asked if price is predictable. I, as a trader, will use any tool i can have access to, to figure out the points where the price movement can be predicted, and i can make a profit.

So while you try to figure out the randomness of the market, I will be driving my demo Ferrari ) trololo
  • #94
  • Dec 15, 2013 12:00pm Dec 15, 2013 12:00pm
Quoting Kanzler
{quote} Independently, candles don't have a lot of value. In context, they can give you a good sign as to whether or not strong buying or selling is going on in a target zone though. It's more about the wicks and quick retreats than anything else.

Yep that shows buying / selling support nicely, major part of how I trade.
Nothing to it, but to do it!!! Stick to the plan FOOL!!!!
  • #95
  • Dec 15, 2013 12:04pm Dec 15, 2013 12:04pm
Quoting BooG
{quote} Ok. So, why are we still here? It means there is just no way to trade the markets ,since they are all a random string of digits and we'll all end up with break even. The thread starter asked if price is predictable. I, as a trader, will use any tool i can have access to, to figure out the points where the price movement can be predicted, and i can make a profit. So while you try to figure out the randomness of the market, I will be driving my demo Ferrari ) trololo

BUT BUT BUT that's not logical at all!!

Listen again, I'm NOT saying the market is random, I'm saying, what you are attempting to use to see the market is RANDOM, therefore using your random methods where you might aswell just flip a coin, is never going to get you your Ferrari, everyone uses your method on here claims to be profitable, but we know it's a lie.

Rule #1 everybody lies until proven otherwise.
Nothing to it, but to do it!!! Stick to the plan FOOL!!!!
  • #96
  • Dec 15, 2013 12:05pm Dec 15, 2013 12:05pm
Quoting Proximus
I can easily demonstrate that its not 50 -50 just by calculating red and green candles on M1 for the past 20 years and you can see that it will be much more candles in 1 type than on the other, way past a reasonable distribution, so in either way its biased.But to say random, i agree.
From my data set:

M1 candles on EUR/USD from 28/11/2001 04:10:00 to 13/12/2013 21:58:00
UP 1734571 41.90%
DOWN 1729001 41.77%
DOJI 675587 16.32%

S&P is upside biaised yet, there are still as many up than down:
M1 candles on S&P500 from 19/08/2008 09:31:00 to 13/12/2013 21:14:00
UP 433274 34.32%
DOWN 423747 33.56
DOJI 405261 32.10

The marginal probability is still 50-50%

----------------------------------------------------------------------------

There is a very good point made my Turveyd about momentum. Market are heteroskedastic (the variance varies over the time).
The variance is the average squared distance of the samples from their mean. The variance is always positive.
When the market has a small variance (volatility) it will have a higher one later. Or it may get a new even smaller one. But it can't have a negative one: it is bounded/blocked. Volatility has to gets bigger at some point. Volatility breakout strategies seems to makes sense.
We can't predict the direction. We can't predict the time. But we can predict that the amplitude will get bigger. OK, we can't predict the magnitude (for a perfect TP) but R:R>1 nonetheless.
Since the market can't stop being heteroskedastic can we consider this as a structural inefficiency of the market?
No greed. No fear. Just maths.
  • #97
  • Dec 15, 2013 12:07pm Dec 15, 2013 12:07pm
Quoting Turveyd
{quote} BUT BUT BUT that's not logical at all!! Listen again, I'm NOT saying the market is random, I'm saying, what you are attempting to use to see the market is RANDOM, therefore using your random methods where you might aswell just flip a coin, is never going to get you your Ferrari, everyone uses your method on here claims to be profitable, but we know it's a lie. Rule #1 everybody lies until proven otherwise.

Turveyd, hi bud long time no see.... This is probably one of the best threads ever started on FF. and I am here by accident lol
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
  • #98
  • Dec 15, 2013 12:12pm Dec 15, 2013 12:12pm
Quoting PipMeUp
{quote} From my data set: M1 candles on EUR/USD from 28/11/2001 04:10:00 to 13/12/2013 21:58:00 UP 1734571 41.90% DOWN 1729001 41.77% DOJI 675587 16.32% S&P is upside biaised yet, there are still as many up than down: M1 candles on S&P500 from 19/08/2008 09:31:00 to 13/12/2013 21:14:00 UP 433274 34.32% DOWN 423747 33.56 DOJI 405261 32.10 The marginal probability is still 50-50% ---------------------------------------------------------------------------- There is a very good point made my Turveyd about momentum. Market are heteroskedastic (the variance...

Now look at how many times M1 bars ( I have candles ) go up in row or down and compare that to H4, you'll get a lot more 10 bar 1 direction moves then H4 charts.

And no Momentum gives you direction, there is then noise within that direction, ie pull backs in a up move which are buying op's.
Nothing to it, but to do it!!! Stick to the plan FOOL!!!!
  • #99
  • Dec 15, 2013 12:12pm Dec 15, 2013 12:12pm
Quoting billytt
{quote} Turveyd, hi bud long time no see.... This is probably one of the best threads ever started on FF. and I am here by accident lol

Yo dude, did I hear you type you've given up on trading ??

Didn't understand why, if so ??
Nothing to it, but to do it!!! Stick to the plan FOOL!!!!
  • #100
  • Dec 15, 2013 12:14pm Dec 15, 2013 12:14pm
Proximus

So in my point of view, an uptrend is nothing more than a biased, already random market, where the bias comes from fundamental factors.
A downtrend is the same but in the other direction, while a sideways movement is the perfect 50-50 situation where there is no bias/trend, so the market will take its real form of 50-50
...you got it, trading whip saw will allow many systems /methods to work and your charts you posted, my favourite as everyone knows was to trade inside bars and open price but that is all history now. One thing ,, the trades might be following the trend but how often will the people be stopped out during this process, its a hard game
SEE LINE,TRADE LINE..PRICE HAS TO GO SOMEWHERE,,, PRICE WILL GO SOMEWHERE.
5 traders viewing now