Cyclical patterns in the market - page 3

New comment
 
Dima.A.:

Whatever...



You know best...
 
prikolnyjkent:

You know best...

explain...
 
Dima.A.:

explain...


Well, firstly, on your chart it's a bit different than on the usual... candlestick chart.

Apparently that's why it makes absolutely no difference whether it's 30 minutes on the chart or what...

 
alsu: ... it just turns out (at least in all the cases I've come across) that the mutual information between the variance processes and the quotes themselves (already taking into account the sign of the increments) is practically zero. In other words, they are simply two multiplicatively superimposed independent random processes. Independent at least so much so, that the problem of practical application is not solved head-on. ...

Thank you.

Somehow I thought that TC checks based on these models did not use directional prediction (conditional variance, etc., no mention of direction)
 
GaryKa:

Thanks.

For some reason I thought that TC checkers based on these models didn't use forecast of direction (conditional variance, etc., no mention of direction).

I've somehow moved away (and advise others) from trying to get a general prediction model for a quote, instead I prefer to identify so to speak "prediction points", where there is local inefficiency, i.e. smearing of information in a limited volume of quotes. I don't even know how to explain it succinctly.... There are such processes, they are called discrete and discontinuous (I quote a picture from the respected Vasily Ivanovich):



So I consider all sorts of archi-garches to be nothing more than phenomenology, describing the process in principle, but not saying a word about its inner workings.

 
prikolnyjkent:


I can suggest a door that leads in that direction. But, the result will be up to you.

Run in the tester (I have ForexTester) a series of trades with a random opening direction and TP=SL=20...25 pips. Open a new position as soon as the existing one is closed. And then - look at the chain of bars, which show completed deals on the chart of the tester (take three days).

If you see something interesting - you will have something to think about. But if you don't - it will be a pity...



Here's what actually happened. I've always noticed that the period of rising period is followed by a falling period in case of random entry and as a result the spread is simply plummeting in the long term, and on the yield chart I can see the fractality, i.e. self-similarity of large parts to small ones. But I'm thinking long and unsuccessfully about how to use it. Is it the right direction? Or am I missing something else?

 
Telo:

"Long and unsuccessful" is just like a forex slogan!
 
Joperniiteatr:
Author, do a discretisation on the tf at least, m1-forecast for how much and what the sky is like, m2-the same and so on

I've read the threads you've got, I've got nothing special to learn from them. I'll make some pictures from different timeframes.
 
alsu:
"Long and unsuccessful" is just like a forex slogan!



do you know how to make money with sb? you are probably too smart for that, and have a huge baggage of knowledge and standard theories which, due to their logicality, do not allow new thoughts to shine through
 
Telo:



This is what actually happened. I have always noticed that the up period is followed by a down period on a random entry and in the long term the spread is simply lost, and on the yield chart you can see the fractality, i.e. self-similarity of large parts to small ones. But I'm thinking long and unsuccessfully about how to use it. Is it the right direction? Or am I missing something else?


You could try to find correlations in the flow of trades, those that are not revealed by the analysis of the prices themselves. In other words, help predict the outcome of a trade using the results of previous trades. But be careful, the perpetrator is armed can reveal what is not really there))
12345678910...20
New comment