why have you accepted the price-predictability assumption? | Page 28 | 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?

  • #541
  • Dec 29, 2013 12:45am Dec 29, 2013 12:45am
Quoting milam
{quote} Demo
So what? This is still a method, better than the best.
The Thief of Wall Street
  • #542
  • Dec 29, 2013 12:51am Dec 29, 2013 12:51am
Chicky, don't be naughty!

that stupid demo you show is a scalping system taking 1 pip profit

1 PIP TP!

are you out of your mind???

you think a normal broker will let that crap run LIVE???

jz

Quoting Chicky
{quote} So what? This is still a method, better than the best.
  • #543
  • Dec 29, 2013 12:55am Dec 29, 2013 12:55am
Quoting Proximus
For example one pattern that has been bothering my @$$ for the past few months is the "U" shaped and inverted "U" pattern.Which is very characteristic to high liquidity and high volatility pairs like the EUR/USD is in some times (like outside LONDON NY session).It generally happens when price is in ranging or consolidation (thus starts in a random enviroment), thus may seem totally random to appear, but it repeats itself too much to be just random even if it appears in a pretty much totally random interval of the market,so must have a background...
For what it might be worth, Id do an enormous amount of research, and there really "IS" something to that pattern. It's a "market maker algorithm" It describes the "Run, Pause" pattern in the Auction Market process, . But the "run, pause" alert is a very tradeable pattern when view from an Auction market perspective. When you expand sections of the MP indicator it looks very similar to the PF chart in the attached example. I've also attached an expanded view of the MP indicator found here at FF. By expanding the profile it allows you to take a closer look at market structure.........

the above reference is found on pg. 7 of the attached "Algo Futures" newsletter
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Capture.JPG Size: 157 KB
Click to Enlarge Name: YMt53vS.gif Size: 32 KB
Attached File(s)
File Type: pdf Algo futures - Trending Now Newsletter Issue 3_20th October.pdf   2.6 MB | 470 downloads
Markets are not efficient, rather they are effective - Jones
  • #544
  • Dec 29, 2013 12:55am Dec 29, 2013 12:55am
Quoting Proximus
The problem is no the laggingness , not at all, the problem is that it simply doesnt work at all, because it reads past data, but how does it read it? We can assume that past data contains future information,pointed out many times in this thread, but its very important how you read the past data.All indicators are static indicators, not dynamic, meaning that they look at past data, but with a given fixed algorithm, which may or may not work.
You need to go back to the start.

The bulk of indicators are based on time and price. These aren't static at all. Apply a constant to a variable, and it will react to the variable, not to the constant.

Instead of looking at why these fail, look at when they succeed.

This will be the first step in the right direction.

There is the option for neural...up to now, anything I have seen has either been pure bullshit, or impractical. Maybe you will be the first.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #545
  • Dec 29, 2013 12:56am Dec 29, 2013 12:56am
Quoting Chicky
{quote} I have no issue you sipping martini on your yacht and get surrounded by beautiful women (provided none of them is any of my wives). But if you really want to do that, code this one .................... 113,000% in less than a month should be enough for any amount to start with {image}
Is that the Lucky EA ?

Quoting charvo
Chicky, don't be naughty! that stupid demo you show is a scalping system taking 1 pip profit 1 PIP TP! are you out of your mind??? you think a normal broker will let that crap run LIVE??? jz {quote}
Agreed.But most of 1 pippers, dont even work in demo if traded with high tick / minute type markets like EUR USD.They exploit the inbetween tick sizes, but if the ticks come too random and too frequent they fail.Not to talk about how slipappage ruins it on REAL.
"There's a sucker born every minute" - P.T. Barnum
  • #546
  • Dec 29, 2013 12:58am Dec 29, 2013 12:58am
Quoting Chicky
{quote} So what? This is still a method, better than the best.
I know you are not serious, anyway…...

In my view today trading is based predominately on mind, a demo account does not apply…...

I once enter a competition with a demo, the best one were making about 700%, I got only tenth with 500% (in a month), do you think I can do that with a live account?
I do not think so. Why? Because mr EGO will come to play..and he will make it very difficult at any chances he got..
Wherever you go, there you are.
  • #547
  • Dec 29, 2013 1:00am Dec 29, 2013 1:00am
Quoting mzvega
{quote} For what it might be worth, Id do an enormous amount of research, and there really "IS" something to that pattern. It's a "market maker algorithm" It describes the "Run, Pause" pattern in the Auction Market process, . But the "run, pause" alert is a very tradeable pattern when view from an Auction market perspective. When you expand sections of the MP indicator it looks very similar to the PF chart in the attached example. I've also attached an expanded view of the MP indicator found here at FF. By expanding the profile it allows you to...

No disrespect meant, I have seen your work in the Auction Market thread...Very Impressive as an alternate perspective...

But

I can do the same thing with an hourly candle applied to the M5 charts with a lot less clutter, and can extend this into daily, weekly and monthly...

Simple hidden gaps created by breakout zones, typical session retraces...
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #548
  • Dec 29, 2013 1:02am Dec 29, 2013 1:02am
Quoting Magix
{quote} No disrespect meant, I have seen your work in the Auction Market thread...Very Impressive as an alternate perspective... But I can do the same thing with an hourly candle applied to the M5 charts with a lot less clutter. Simple hidden gaps created by breakout zones, typical session retraces...
None taken!

Good Luck with your efforts!
Markets are not efficient, rather they are effective - Jones
  • #549
  • Dec 29, 2013 1:03am Dec 29, 2013 1:03am
Quoting milam
{quote} I know you are not serious, anyway…... In my view today trading is based predominately on mind, a demo account does not apply…... I once enter a competition with a demo, the best one were making about 700%, I got only tenth with 500% (in a month), do you think I can do that with a live account? I do not think so. Why? Because mr EGO will come to play..and he will make it very difficult to at any chances he got..
That's your choice. I make 100,000% a month using Japan's Tihs Si Taht brokers.
The Thief of Wall Street
  • #550
  • Dec 29, 2013 1:04am Dec 29, 2013 1:04am
Quoting Magix
{quote} You need to go back to the start. The bulk of indicators are based on time and price. These aren't static at all. Apply a constant to a variable, and it will react to the variable, not to the constant. Instead of looking at why these fail, look at when they succeed. This will be the first step in the right direction. There is the option for neural...up to now, anything I have seen has either been pure bullshit, or impractical. Maybe you will be the first.
No-no, the price is dynamic, the indicator is not. The price changes while the indicator applies the same code/algoritm to it while disregarding suddent changes, corrections or structure formations in the price.So only the market changes while the indicator does the same.

The stochastic is just using a bunch of moving averages and ranges them between 2 points, the fact that they try to predict the market is no more than pure chance, they dont predict the market, they just run an algo on it, and they either hit it right or miss it, thats why they are not accurate at all.

A true dynamic indi would be one which has an initial value set by careful research, then when run on a forward test it learns the new pattenrs, adapts to price change, and can memorize structure changes.So a neural network could be the closest thing to a dynamic indicator i can think of.

Stuff like MACD , stoch and other crap are just a bunch of averages prices and a few calculations done with them, while disregarding the main adapting feature.
"There's a sucker born every minute" - P.T. Barnum
  • #551
  • Dec 29, 2013 1:06am Dec 29, 2013 1:06am
Quoting mzvega
{quote} None taken! Good Luck with your efforts!

Admittedly...I still do browse in to see what you have posted as confirmation...



Keep on Rocking it!!

Your work and effort are a credit to FF and others that can follow you.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #552
  • Dec 29, 2013 1:09am Dec 29, 2013 1:09am
Quoting Proximus
{quote} No-no, the price is dynamic, the indicator is not. The price changes while the indicator applies the same code/algoritm to it while disregarding suddent changes, corrections or structure formations in the price.So only the market changes while the indicator does the same. The stochastic is just using a bunch of moving averages and ranges them between 2 points, the fact that they try to predict the market is no more than pure chance, they dont predict the market, they just run an algo on it, and they either hit it right or miss it, thats...
Yup.

You need a lot more application, and a lot less information.

Price in relation to Price in relation to Time...

DYNAMIC.

They are not meant to forecast where price will head, they will provide entries where long trends are oversold and short trends are overbought.

Change your perspective, you're viewing information with tainted optics.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #553
  • Dec 29, 2013 1:20am Dec 29, 2013 1:20am
Quoting Magix
{quote} Yup. You need a lot more application, and a lot less information. Price in relation to Price in relation to Time... DYNAMIC. They are not meant to forecast where price will head, they will provide entries where long trends are oversold and short trends are overbought. Change your perspective, you're viewing information with tainted optics.
I understand what you are saying but do you understand me?

I know that stoch is basically still the price in some form, but the dynamicness and adaptative ability doesnt come from the price itself.It comes from memorizing past situations and correcting errors.

If i run a stochastic it wont give me a better edge then looking at the naked price alone and trying to guess the direction.What it would be the best is an algorithm which changes also, not only the input.

Thus the input is dynamic of course the market changes, but so does the filters and conditions have to, for example if you'd traded x pattern with a 150 pip daily average volatility for a few months and in the past weeks now only works with an average volatility of 140 pips, the algorithms must know to adapt to that and replace the old filter.

This is what the neural network does, input comes in, but the threshold and the filters are dynamic and ever changing and adapting to the current conditions.
The NN would be the Holy Grail if they are set up propertly, which again would take many many time of your life and a very fast pc to compute that crap.This is why most NN builders fail because they have limited resource.If a science laboratory has limited breakthoughs using NN to predict price action, what chance does an average trader like us have?

But that doesnt mean that NN are crap, it only means that we havent discovered the best one, and i agree that it would be impractical to waste your time with them, there are better holy grails out there
"There's a sucker born every minute" - P.T. Barnum
  • #554
  • Dec 29, 2013 1:23am Dec 29, 2013 1:23am
Quoting Chicky
{quote} That's your choice. I make 100,000% a month using Japan's Tihs Si Taht brokers.
lol

You know I think many traders and specially on FF fail because they have too much expectation….
I know somebody that if you show him 20% return a year, he will find money for you to trade (6 month books minimum is needed). He pays 15% of your return.
If he allocates 10,000,000(which is not really much for those people) to the account, a 20% will return two million, the 15% of that is 300,000. Not bad I think….
Wherever you go, there you are.
  • #555
  • Dec 29, 2013 1:30am Dec 29, 2013 1:30am
Quoting Proximus
This is what the neural network does, input comes in, but the threshold and the filters are dynamic and ever changing and adapting to the current conditions.

I understand what you are saying...

But why would an NN be any different. It still has to look at the most recent data in order to adapt to the ranges...ie, it is still working on historical data.

Yes, these ranges can change from day to week to month to year...appreciated, and I have even commented on the same thing for adaptation of these ranges. Take an average day that trades a little out of bounds and it skews the rest. Additionally, taking Sunday Data into a daily consideration for a 5 day trading week throws the entire week out...(yes I know this can be scrubbed, just an example)...

Moving from here, there are session ranges and News Exceptions. FOMC that drives a 400 pip GBPUSD market outside of London session with no retrace is hard to adapt to in a down trending market. All things that need be taken into consideration.

Moving averages are a lot more true as they adapt to price, however can lag in a whippy market as well as the price in relation to these in the CCI.

Question is, do you trade by exception, or do you trade by rule?

If you are shooting for 100% trigger win ratios, NNs too fail. No matter which logic you trade by, it's always going to boil down to when to cut losses and when to take profits.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #556
  • Dec 29, 2013 1:33am Dec 29, 2013 1:33am
Quoting milam
{quote} lol You know I think many traders and specially on FF fail because they have too much expectation…. I know somebody that if you show him 20% return a year, he will find money for you to trade (6 month books minimum is needed). He pays 15% of your return. If he allocates 10,000,000(which is not really much for those people) to the account, a 20% will return two million, the 15% of that is 300,000. Not bad I think….
I was granted my first live Managed Account based on a 1 week trial with 1.8% net gains.

Was a $10k account and I was ashamed of the gains...I was given 60,000 Euro to trade the following week.

You are correct with your statement on the expectation of others.

How many think they can start with $1000 and make a living?

Sure it can be done, but consistently and for how long? The pressure behind this is more than the typical person can handle.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #557
  • Dec 29, 2013 1:34am Dec 29, 2013 1:34am
Quoting Chicky
{quote} That's your choice. I make 100,000% a month using Japan's Tihs Si Taht brokers.
Yeah sure that is why your signature shows a -97% return.

If you'd made that much you would not be sitting here and talking with us.If you'd even start with 100$ and made a 100.000% return monthly then in only 4 months trading you could aquire more money than it is currently on the entire planet, numerically: 100,501,001,000,500,100 thats 100 quadrillion bucks .
"There's a sucker born every minute" - P.T. Barnum
  • #558
  • Dec 29, 2013 1:37am Dec 29, 2013 1:37am
Quoting Proximus
{quote} Yeah sure that is why your signature shows a -97% return. If you'd made that much you would not be sitting here and talking with us.If you'd even start with 100$ and made a 100.000% return monthly then in only 4 months trading you could aquire more money than it is currently on the entire planet, numerically: 100,501,001,000,500,100 thats 100 quadrillion bucks .
Do you really believe there was a need to do this calculation and write a reply to my above post? No wonder you cannot write an EA.

ps. -97% is far better than +130,000%, ask me why?
The Thief of Wall Street
  • #559
  • Dec 29, 2013 1:46am Dec 29, 2013 1:46am
Quoting Chicky
{quote} Do you really believe there was a need to do this calculation and write a reply to my above post? No wonder you cannot write an EA. ps. -97% is far better than +130,000%, ask me why?
Who said i cannot write an EA.Besides i`m not interested in 1 pippers, ruled those out of my book about a year ago.Most of them cant even stand on my high quality historical data, let alone stand a chance in a demo or even a real account.

They could only and only work if spread = 0 and slippage is the most minimum = you have to be the bank not the trader

I dont know i would still take the 130k % increase anyway if i could , but i only need 1 months like that, i dont care if i never make a dime after that again, just gimme 1 month that result and i can live happy after.
"There's a sucker born every minute" - P.T. Barnum
  • #560
  • Dec 29, 2013 1:47am Dec 29, 2013 1:47am
Quoting Magix
{quote} I was granted my first live Managed Account based on a 1 week trial with 1.8% net gains. Was a $10k account and I was ashamed of the gains...I was given 60,000 Euro to trade the following week. You are correct with your statement on the expectation of others. How many think they can start with $1000 and make a living? Sure it can be done, but consistently and for how long? The pressure behind this is more than the typical person can handle.
I once tried to make $500 a month starting with $5,000. The pressure was way too much than one could handle. I reduced this to $60 a month with $200, even that was very stressful. Then someone turned on the light of the dark room I was sitting in. I now don't expect a regular return, just keep on taking small risk, small reward trades whenever possible. Like Soros (or whoever) said 'walking over 10 of 1 ft. bars than trying to jump over one 10ft bar.'
The Thief of Wall Street
1 trader viewing now