Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Optimized Risk vs Reward Equation 335 replies

Loss Less Optimized Strategy 63 replies

Optimized Parameters in Metatrader Testing? 3 replies

"Optimized Parameter" feature in EA? 0 replies

Strategy with Optimized parameter 1 reply

Attachments: Optimized Trend Trading
Exit Attachments

Optimized Trend Trading

Quoting Jack168
Hi, John,

I don't use shift. Window size, I use 100 to 170 bars.

Colour candles, GRaB Candle Long, GRaB_indicators.rar,... something or, stoch_candle, RSI_candle indicator...

Good weekend.
Thank you and have a nice week-end
Hello all,

I re-read the thread and i see somewhere that John speaks about the :"KH i-Regr v1.mq4 ".

But i don't find it in the attachments section ?

If anyone have this, can you share it please ?

Thanks !
Quoting crashout
Hello all,
I re-read the thread and i see somewhere that John speaks about the :"KH i-Regr v1.mq4 ". But i don't find it in the attachments section ?
If anyone have this, can you share it please ?
Thanks !
Well, I found this release in my magician's hat.

The smoothing is very interesting... but ... but...all those indicators work w/ the concept of constant weights attached at each values (prices and volume), like if any independant value were not separable from others, and if any independant value got the same weight as all other values. worse, there is no concept of market phase distinction.

Remind, all values shown on markets are biaised and should be understood as estimated values shown on the market. The reality of the market can not be drawn on a chart.

Rgds

Did you know ? Apple has more cash than the US government !
Attached File
File Type: mq4 KH NLev-Regr CurveFit v3.mq4   13 KB | 2,012 downloads
Quoting jaguar1637
Well, I found this release in my magician's hat.

The smoothing is very interesting... but ... but...all those indicators work w/ the concept of constant weights attached at each values (prices and volume), like if any independant value were not separable from others, and if any independant value got the same weight as all other values. worse, there is no concept of market phase distinction.

Remind, all values shown on markets are biaised and should be understood as estimated values shown on the market. The reality of the market can not be...

Thanks for the indicator, very nice to share !

For apple yes, i read that on business insider i think.

76 billions, come on Steve, we just need some free macbook pro
Quoting crashout
Thanks for the indicator, very nice to share !

For apple yes, i read that on business insider i think.

76 billions, come on Steve, we just need some free macbook pro
Yep, note I have modified some parameters for optimization, in the release I just sent to you

extern int RegrOrdr = 5; // 3; degree (5 seemed better)
extern double StdDev = 2; // Kernel Standard Deviation
extern int RegrPrd = 21;
extern int AppPrice = PRICECLS;

RegrOrdr should be at least 3, or 5 or 7 (always impair)
StdDev 2 ( may be 2.2 or 2.4) I tried several values for testing
RegrPrd 21 ( I tried other values , you can try 17 or 19)
extern int AppPrice = PRICECLS; (I tried PRICEHA w/o being successfull)

I decreased the number of bars from 10000 to 3000,

I tried to implement the Polynomial Regression on the StochasticDiNapoli, but did not finish it, (lack of time)

Yep, Apple is richer than the US government !
I have found two kinds of Hull Indicators.
This is the result of the combination.
The parameters for HMA Color : 20 - 1 - 0

These for HMA Russian Color : 21 - 3 - 0

I tri to do better but in vain
Attached Image (click to enlarge)
Click to Enlarge Name: eurusd-hull.gif Size: 20 KB
Attached Files
File Type: mq4 HMA_Color_nrp_.mq4   4 KB | 1,394 downloads
File Type: mq4 HMA_Russian_Color.mq4   4 KB | 1,333 downloads
Quoting balouf
I have found two kinds of Hull Indicators.
This is the result of the combination.
The parameters for HMA Color : 20 - 1 - 0

These for HMA Russian Color : 21 - 3 - 0

I tri to do better but in vain
HMA repainting ???

if yes, forget it
HI
There are several way tos deal w/ time-series.
neural stuff and pattern searches are common tools

I tried several and different things regarding algorythms based on regression linear. So, calculating shows me that unit Root testing is motivated by the need to avoid spurious regressions.

Also, after playing w/ Burg alogrythms, I notice regression coefficient seemed to be lagged

The usual t distribution is not applicable but simulation is needed to construct a suitable sampling distribution of test stat.
Also, we are in front of the well-known problem in econometrics that errors are autocorrelated.

After that, I reach the point of the Cointegration engle-Granger . IMHO, the only way to implement this, is to use R.

Well, is there anybody able to answer me regarding the possibility ?

The goals are simple:
- fetch a time -series
- apply the cointegration engle-granger
- launch a regression w/ this algo

rgds
Hi, John,

I haven't read this article yet. Two men did get Nobal prize on this topic, I believe. Google key word, may find the paper.

I found 2 problems are hard to understand or fix.
1. when a nonlinear regression (Quad) just establish(find) a possible trend, then, price smash the band very fast. In forex, "V" type of reversal often happen.

2. price(bar) often move out of regression band, and became data island, and it cause regression result to be changed swiftly, 1 or 2 data can change regression result to be completely different. Terrible problem !

Cheers!
Attached Image (click to enlarge)
Click to Enlarge Name: ProblemforReg_euro_M15_12_8_2011.jpg Size: 95 KB
Quoting Jack168
I found 2 problems are hard to understand or fix.
1. when a nonlinear regression (Quad) just establish(find) a possible trend, then, price smash the band very fast. In forex, "V" type of reversal often happen.
2. price(bar) often move out of regression band, and became data island, and it cause regression result to be changed swiftly, 1 or 2 data can change regression result to be completely different. Terrible problem !

Cheers!
So now, is there any tool to watch accumulation orders ?

This is a mandatory one

Rgds
john
Quoting jaguar1637
So now, is there any tool to watch accumulation orders ?

This is a mandatory one

Rgds
john
Jaguar,

Maybe www.orderbookfx.com !?
Quoting jaguar1637
So now, is there any tool to watch accumulation orders ?

This is a mandatory one

Rgds
john
I am not sure about "accumulation orders" ? what do you means? dealer's level 2 order book?

There is tool to monitor the state of profit/loss of all open trades(open orders). You can google to find out in ff, may be in EA: snow_rolls. That guy post an indicator to show all win/loss of open trades.

My EA do a calculation for state of all open trades after a new bar processing, and store equity and balance values into global variable in mt4. Then, an indicator use data to shows equity curve in a another window.

Try to do an experiment about track equity curve(win/loss of all open trades) for position management. I found, it is a new issue to protect the equity curve for volatile fx trading, which may be a key issue for optimized trend trading. No much talking on this point from guru or expert. Don't know why?
Quoting surfeur
Jaguar,

Maybe www.orderbookfx.com !?
Thank you

I will check it

Rgds john
Quoting Jack168
I am not sure about "accumulation orders" ? what do you means? dealer's level 2 order book?

There is tool to monitor the state of profit/loss of all open trades(open orders). You can google to find out in ff, may be in EA: snow_rolls. That guy post an indicator to show all win/loss of open trades.

My EA do a calculation for state of all open trades after a new bar processing, and store equity and balance values into global variable in mt4. Then, an indicator use data to shows equity curve in a another window.

Try to do an experiment...
Yep!
Can you help me ?
I find nothing w/ google about snow_rolls

In fact, I am asking to watch all open trades on the currency chart, not just the ones opened by one EA on a personal account

Rgds
Hi, John,

Here is the threat on ff:

http://www.forexfactory.com/showthread.php?t=286352

He call Basket_Stats, mt4 indicators are in his webpage.


there are other mt4 indicator: i-Profittrader; iExposure.
Quoting Jack168
Hi, John,

Here is the threat on ff:

http://www.forexfactory.com/showthread.php?t=286352

He call Basket_Stats, mt4 indicators are in his webpage.
there are other mt4 indicator: i-Profittrader; iExposure.
thank a lot !
Hi Guys

For those who are afraid about prediction tolls, take a look on this:

http://www.smartplanet.com/blog/scie...on-models/9840

This tool shoudl work on FF, isn'it ??

Rgds john
Hi, John,

There are plenty of tools for forex prediction(forecast), some are open, some are unknown yet.

Data-driven alpha models are used by somebody, but, no news about this area. Data mining is a wide research area for while. I am not expert in this area, will find time to study.

This book "Evidence-Based Technical Analysis: Applying the Scientific Method and Statistical Inference to Trading Signals" by David Aronson
may be a start point to enroute into this area. I claim I haven't read this book yet. I will have to read it in sometime. Any opinion about book?

Hi, John,

There is a research article about "Non-Linear Predictability" from FED.
But,
1. I don't know if TAR or STAR is the same with your i-Regr?
2. Can anybody create TAR or STAR in mt4 code? Then, we can try the method in fx to see if it is workable?
Attached File
File Type: pdf Non-Linear Predictability in Stock and Bond Returns WORKING PAPER2008-010.pdf   1.2 MB | 1,256 downloads
Hello Trader.

I write my master thesis about algo trading.
And I try to figure out, If it’s necessary or not for a Property Trader (a trader how trade only with his own money) to use algo trading.

What is your opinion and experience???
Thx
0 traders viewing now