Cycle analysis

Good morning,

I am new and this is my first post.

With the help of a friend we've developed in Matlab a cyclical approach to model and predict market behavior, from a scientific point of view.

We got very nice results, and we think we can get even more with an in-depth study and analysis of every step of the whole process.

For this reason we are seeking for people with good background in mathematics to create a study group about cycles and periodicities in financial markets; engineers, mathematicians, physicists etc. are all welcome.

The attach is a sample of what has been done so far, of course if we always get such results we do not need to study anymore...

If there is somebody interested in, please contact me directly via Skype, my nick is acepsut

Greetings from Italy and apologize for my bad english.

Alberto

Interesting work.

But how often do you get those results? Was this a lucky hit/miss example? Or can you consistently (within let's say 0.1 standard deviation) hit the target?

nice...

Well, I know, this is a trading system section and our study is not a trading system, or, better, is not a trading system as of now.

Yes the image I uploaded show a successfully tracking of out of sample data, and honestly I must say we have wrong predictions, but I always keep in mind that this kind of analysis is not a up/down, yes or no output rather than a complex composite blend of directions, slopes, and turning points matches.

I hardly believe such results come from casualities.

We've developed a code that often show incredible results, I can't say how many times it tracks faithfully 30, 40 or even 60 points aheads, but it does!

And we know there is a lot to do to improve what's done; we didn't dive in details of every step and this is exacly what we have plan to do.

Once the Cyclope will be fully developed in every details then we can measure its performance as hit/miss, equity and so on.

For now we are seeking for anybody with good maths skills, some free time and willing to work togheter.

Regards,

Alberto

I am by no means a math genius. But surely I can help in other avenues. (MQL programming)

Ronald you are welcome, just add me to your Skype contact list and let's see if we can start working togheter!

Alberto

Well, since this is a forex forum get a quick look at how cycle code behave in different market, this is a sample from eur/usd 15 minutes time frame.

I think quality predict speaks for itself.

As a reminder, the "in sample" data (known data) ends at point 60, and blue line is forecast. Red bars are later overplotted to verify forecast accuracy.

Hello,

I'm no genius myself but i am pretty god with numbers I can probably be some help if you are in need!!!

Hello Rbi,

you are welcome! I look forward to see you on Skype to join us and for any questions.

Regards,

Alberto

Alberto,

is this your website ?

http://www.advancedsourcecode.com/ne...orecasting.asp

Sixer

Hello Sixer,

I have no website, no blog and nothing to sell, so I have nothing to do with this link or any other.

Regards,

Alberto

Alberto,

thanks for the clarification - I have a Matlab student version. Is it possible to do the shown forecasts by using this basic version or do I need any update ?
My programming knowledge is "zero".

Sixer

Yes I think the student version would work correctly since we've written all code manually, in all steps and all functions, so for now there is no need for an higher version.

The problem may be in - My programming knowledge is "zero"- since we need mathematicians, engineers and so on with good programming knowledge!


Regards,

Alberto

Alberto,

I´m ready to realize the backtesting activities if there is any demand. I have approx. 15 years experience in the stock market. Send me a BM if you need help.

Sixer

Hi Alberto,

Are you still working on cycle analysis ? I've been experimenting with cycle analysis. I use a commercial software Timing Solution for this purpose. Here's my view on cycle analysis:

1) Cycle analysis cannot be used as an independent trade signal. You can only use it to confirm trade signal from other systems/analysis.

2) Cycle analysis doesn't work well in a market with strong trend. However, it doesn't work well in a ranging market with narrow range either.

Just my two cents above. Any comments ?

Hi pooh,

as far as I know, Timing Solution use a spectral analysis method (don't know which one) to exctract cycles then auto-select and keep only the highest amplitudes ones.

There are different methods and approaches to perform cycle analysis, and in my opinion this is not the best one. For what about your questions, I think cycle analysis can be used as a standolone tool if and only it is well designed and applied.

My researches shows that this is a very powerful tool, but the main problem for me is about when to use it and when not to use it. It is a little bit like Elliot wave analysis, you can't use it anywhere anytime, but when a movement clearly unfold, it is a very good predictive tool.

You said it doesn't work in strong trend and narrow range sideway market: that's right because a very strong trend market acts like a straight line with a steep slope, and this can be very difficult to model. In cycle analysis this is seen as a very low frequency driving the market, while in a narrow sideway range there are only high freq present, both of this usually are discarded in pre processing phase.

This brings back to what I said previously, the main problem is to be sure (or at least, very confident) about when to use and when not to use cycle analysis.

Hello; your English is okay! By the way, according to my personal trading experience; there is no ending point of learning! So, don’t stop your practice and strategy update session even though you are making regular money in your strategy’ it’s a decentralized market.