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

Hello all,


I just stumbled on this thread, and it sadly doesn't appear to be very active any longer.

I have an educational and professional background in math, and prefer to use math/stats concepts over classic technical analysis for generating trading signals.

I use stat arb methods, as well as numerical approaches such as GARCH, Kalman/Weiner filters, and finite difference methods. I'm also interested in machine learning techniques.

Someone asked a question about GARCH. On another computer (the one that is running my self-designed automated strategy), I have a GARCH routine that I coded in MQL4. I'm happy to share it if you can wait a few days.

I would be interested in discussing research ideas or specific techniques. What I'm not interested in is using off-the-shelf EAs or pure indicator-based systems.
Some time ago there was a paper posted in this thread on market microstructure, including a simulation of order book, etc.

I can't find that paper now, does anyone have a reference to it?
Quoting flyboy
Some time ago there was a paper posted in this thread on market microstructure, including a simulation of order book, etc.

I can't find that paper now, does anyone have a reference to it?

You can google Richard K. Lyons and Carol L. Osler, get into their webpage. These two prof are leaders in fx microstructure research.

I attach their papers as follows. I haven't studied their papers yet and their papers are bit of old, I believe.
Attached Files
File Type: pdf New Micro by Lyons.pdf   524 KB | 937 downloads
File Type: pdf FXMicrostructure-Survey_07-30-08.pdf   663 KB | 1,016 download
Quoting oblomov
Hello all,


I just stumbled on this thread, and it sadly doesn't appear to be very active any longer.

I have an educational and professional background in math, and prefer to use math/stats concepts over classic technical analysis for generating trading signals.

I use stat arb methods, as well as numerical approaches such as GARCH, Kalman/Weiner filters, and finite difference methods. I'm also interested in machine learning techniques.

Someone asked a question about GARCH. On another computer (the one that is running my self-designed automated...

Hello, Oblomov,

I am interested in Garch model, and collect some papers, and wait my time to do deep study and code. It seems no mt4 code for Garch. ok, waiting for you to test your mt4 code.

these are good papers:
1. "Bayesian inference on GARCH models using the Gibbs
sampler" by LUC BAUWENS, MICHEL LUBRANO.

2. "Estimation and Prediction of Generalized Growth Curve

with Grouping Variances in AR(q) Dependence Structure"
by Jack C. Lee and Ying-Lin Hsu
If you like math and coding from math, then 2nd paper is quite good and significant method ---- my favourable theory of fx trading ----

Due to All forecast are based on growth and growth curve. FX trading is a speculation of the growth on some fundamental economics factors.


Cheers !
Attached Files
File Type: pdf Bayesian inference on GARCH models using the Gibbs sampler.pdf   859 KB | 903 downloads
File Type: pdf Estimation and Prediction of Generalized Growth Curve.pdf   160 KB | 815 downloads
I am still looking for my code that shows the GARCH variance estimates as indicators.

There are similar models, like CARR, that researchers have found to be superior to GARCH. The attached paper reviews the key models.

The key insight from this paper and the other research on volatility is that range-based volatility models (ones that use the OHLC values instead of the classic log returns as the definition of volatility) are superior to returns-based volatility models.

Another way to put this is that Welles Wilder had the right idea with ATR. However, ATR is not an efficient estimator compared to ones such as Yang-Zhang. To his credit, Wilder was ahead of the leading academic researchers on this by 20 years.
Attached File
File Type: pdf 1-4 Nathan Liu_Range_Review_NCTU.pdf   151 KB | 1,339 download
I just got a link it is about The GARCH Model in mt5. I hope that helps.
http://www.mql5.com/en/articles/222
Quoting ChesterB
Hello John and all. I recently stumbled into this thread, cycles, and Ehlers... I initially started trading years ago, stopped, and just started again in the last several months.

My recent results, although overall profitable, have been very erratic, and my methods have been discretionary.

Intuitively, I have been trading cycles, but without a clear system.

Usually this is on the 15M timeframe...

I watch price go up , say 25 pips over 8-10 bars, go back down 25 pips over the same amount of bars, place a long trade in anticipation...
Hi ChesterB!

Did you ever make any progress in systemizing this?
Curious to learn what you came up with.
Best,
Cam
Quoting John Last
I want you to show shots of another two mods. One is Brain Trend of SSA.
You can do it yourself. All you have to do is to replace the jjma in the iCustom function. Use lags of 50 bars with 5 number of computations.

Unfortunately this mod relies on an indicator of Mladen SSA of price, which is in the Elite section of TSS. So I can't share in a public forum. But hey I tell you what to do.

The other reason I will not release that is that this is expected to recalculate. How that will handle I do not know.

You can look at the BB MACD with...
Hello, i'm a new guy on this forum, i tested your indicators but, thei are don't works, i want to know why ? please because i'm interested by all your indicators.
Hi,

It is a problem with your broker mt4. If you use another demo they will work.
Anyway check the latest versions here:

http://beathespread.com/file/view/291/brain-trend-club-files

Tu es bienvenu sur le site.
Very please with it my email [email protected]


thank you.
Attached Image (click to enlarge)
Click to Enlarge Name: e-u.jpg Size: 103 KB
Hi,

I noticed that there was afew posts on discussion of Hilbert Huang Transform (HHT). Has anyone coded it out yet for MT4? The Russians have coded out a version but sadly its for MT5. Kindly post here if you do know of any MT4 version. Thanks

http://www.mql5.com/en/articles/439
Quoting mefesis
Very please with it my email [email protected]

http://www.forexfactory.com/attachme...2&d=1348326368

thank you.
Hi do u have this indicators on the main screen if yes pls post the link.

Thanks in advance
Quoting oblomov
Hello all,


I just stumbled on this thread, and it sadly doesn't appear to be very active any longer.

I have an educational and professional background in math, and prefer to use math/stats concepts over classic technical analysis for generating trading signals.

I use stat arb methods, as well as numerical approaches such as GARCH, Kalman/Weiner filters, and finite difference methods. I'm also interested in machine learning techniques.

Someone asked a question about GARCH. On another computer (the one that is running my self-designed...
looking forward to your garch indi!that's will be great.
garch with R is really take a lot of time...
Quoting mefesis
Very please with it my email [email protected]


thank you.
Please send me, my email [email protected]
Quoting nguyentho
Please send me, my email [email protected]
can you please forward to my email too? [email protected]
Hudson Metrics
Quoting quantq3
can you please forward to my email too? [email protected]
Really, if your email address is Quantechtrading, that means you work as a member of this financial company (http://www.quantechtrading.com ), I can not believe and worse, do not understand why you are in need of this kind of indicators... may be at your job, they did not give you all their knowledges....

And if, as they said, "Success only comes from Strength. Strength of Vision. Strength of Character. Strength of an Idea.", sure, you should have much better in your baggrab than us

So, share your indicators you are working with, this will be a benefit for the public.

Jaguar
hi new to forex factory.. how did codebreakers trend trader go, i didnt read through the entire post only the first page.. was it as good as he stated?
I am interested in working with a couple of people to build a predictive time series forecast process based on cycles. I have some clever ideas to go forward with that are not well known and would appreciate the help of a coder, maths orientated person and thats about it.

I have a mathematical was to evaluate cycles, and in effect filter the dominant cycles. I think that my ideas are a better representation of cycle strength and phase then is currently understood. The mistake I think you are all making is that the current classification of cycles simply does not function in the financial markets and that your ideas of cycles need to be challenged, because they do not work on a reliable basis.

Please PM me if interested.


Mike Haran

As I was having a shower I was thinking about cycles as you do, I realised that if you are using metatrader data to analyse cycles you will ALWAYS be wrong. Market cycles are based on belief, a very human emotion, human emotion is not turned off at the weekend or over holiday periods, we continue to think, worry, and plan regardless of whether the market is closed or open. These ideas do not switch off they continue to form and evolve. No matter how good your market spectral analysis software is you WILL fail to predict highs and lows. The earth continues to spin while we are asleep, you wake up and its morning, just because you were asleep the market still continued to move.

Imagine you are happy, how happy can you be? At some point you can be no happier, the limit of your emotion is reached, no matter what good things happen you are fully extended, there can be no more additional pleasure. The same is with fear, at some point the limits of your mind to experience fear is fully saturated, you can take no more, you collapse or die of fright. So it is with traders, at a point they can take no more and they abandon their trades in fear or elation, and the market turns.
Just to prove a point the GBPJPY should make a top somewhere about the 23-31th Jan and then fall about 500 points. My cycle analysis would point to 145 area around the 23rd Jan or 31st Jan. I will now estimate the correction length and time.

Mike Haran
Gbp-Jpy, H4, next cycle low (trough) may be at 20 gmt, 23 January. If next low is a lower low, then, afterward peak may be a lower high.

Cycle prediction got a problem: you may get a forecast of period or time correct, but, price may be not a high peak or that cycle high/low got changed.

Market is often in cycle move, but, period of cycle is often changed by news event, price spike, big guy change their mode, etc...

I found Casino Baccarat game have cycle sometimes, banker and player have their own cycle. But, when we win on this cycle method, the house (dealer) can use shuffle machine (or pre-shuffled cards) to build different cycle or remove cycle off this shoe. They can cheat you, and beat you out.
0 traders viewing now