For those who have (are) seriously engaged in co-movement analysis of financial instruments (> 2) - page 20

New comment
 

I understand a lot of people want to speak out. And the thread is starting to get a bit of a flood...

I draw attention to the title of the thread.

 
I understand too, you can be interested in different ways, some just theoretically, some practically. You should at least post research results (like: such-and-such pairs with such-and-such coefficients give a channel with such-and-such duration (it is not obligatory)), because everything comes down to searching for coefficients (because the number of instruments in different DCs is limited) to build a channel synthetic with the help of your recirculation.
 

The thread showed that either there are no people on this forum, or they don't want to discuss the topic indicated in the title.

Recycle refers to the topic in question. But the topic is much broader than that. The only constructive work on the subject is presented by 7bit. Nothing else.

All the rest is just hot air, unfortunately.

I already told you that I will not comment on Recycle in this thread anymore. I have voiced and demonstrated my vision of correct approaches here many times.

I would like to see a discussion of the topic, not a "jackass".

 
hrenfx:

The thread showed that either there are no people on this forum, or they don't want to discuss the topic indicated in the title.

Recycle refers to the topic in question. But the topic is much broader than that. The only constructive work on the subject is presented by 7bit. Nothing else.

All the rest is just hot air, unfortunately.

I already told you that I will not comment on Recycle in this thread anymore. I have voiced and demonstrated my vision of correct approaches here many times.

I would like to see a discussion of the subject, and not an "asshole himself".

), i.e. you have not offered anything constructive either? only 7bit on the horse. )

ZS: couldn't find his posts in this thread. )

 

Constructive criticism of 7bit's work with justification has been made. I have no intention of PRing my own work.

If there is criticism, it should be justified. If there is a suggestion, it should also be justified.

 
hrenfx:

Constructive criticism of 7bit's work with justification has been made. I have no intention of PRing my own work.

If there is criticism, it should be justified. If there is a suggestion, it should also be justified.

I would have criticized, but I am drunk (afraid to be rude), in the subject not really understand, but the essence is understood, more understandable to the people, you may folk craftsmen and would pull up. And so unpretentious mathematical arrogance (sorry, of course).
 
hrenfx:

All the rest is just air talk, unfortunately.

I have already said that I will not be making any more comments on Recycle in this thread. I have voiced and shown my vision of correct approaches here many times.

I would like to see a discussion of the subject and not an "asshole himself".

Discussing someone else's vision without any specifics (cranes in the sky) is a lot of fuss. There is nothing to discuss except the author of the recycle: whether he is a jerk himself or someone helped him. Therefore, in order to have a concrete discussion, you must first present concrete results for public scrutiny:

sanyooooook:

...

You should at least post results of your research (like: such-and-such pairs with such-and-such give a channel with such-and-such durations).

 
Reshetov:

Are they too lazy or what? The only person who was not lazy and looked it up was genro.

Everything has been laid out, told and explained on my part. You have not started even once with default settings. Do I have to do it for you? (You don't have to answer).

 
hrenfx:

1) The thread showed that either there are no people on this forum, or they don't want to discuss the topic indicated in the title.

Recycle refers to the topic in question. But the topic is much broader than that. The only constructive work on the topic is presented by 7bit. Nothing else.

2) All the rest is just hot air, unfortunately.

I have already said that I will not comment on Recycle. I have voiced and demonstrated my vision of correct approaches here more than once.

3) I would like to see a discussion of the subject and not an "asshole himself".

1) I honestly admit - I am greedy. I have ideas but I don't want to post them on the forum.

2) I agree. I mostly develop willpower, learn to read the empty stuff and keep quiet with my corrections. So, sometimes I just parrot it for the fun of it. Nothing serious. ;)

--

3) OK. I'll give you some ideas. They're abstract enough to make the freeloaders go berserk.

First of all.

In all variants (of which there are two basic poles: trend one (trendomat, Yury's portfolios) and counter-trend one (arbomat, recycle))

Some known forecast is implicitly used. I.e. there is an attempt to "drive" the synthetic into either "purely trend" or "purely flat" when composing the portfolio,

and then have it based on one of two fabulous assumptions 1) "If there is a trend, it will continue" 2) "Price always comes back".

Where I'm going with this is that it all seems a bit absurd to me (well, to be honest, just unprofitable). Why? Quotes themselves

don't tend to stay in one tale for long, they like variety.

From this I draw (made long ago) a simple conclusion: "You should not fight the price, putting it into your forecast. It doesn't like that and will always commit

Trying to escape from the Procrustean channel. Nada love it as it is, and that is how it is forecasted." . First you have to understand this. And then...

you can use multicurrency patterns.

Secondly.

It follows harmoniously from "in the first place". If in the first point there is a slightest progress, then even a simple circular multicurrency connectivity of Forex allows us to solve this problem.

The problem is to turn it into a ready-made (well, almost ready-made) free neuronet. Of course, we still need to figure out how to do it, but that's exactly the topic I will not discuss here. ;)

--

Finally I'll tell you the commonplace truth - statistics won't work as oracle, no matter how you twist it. No great R will not help.

When sorted by boxes, shelves and ranks, the dynamics of the original sequence is lost, and the basis for a full-fledged

(read high-yield) forecasting. That leaves statistical stumps and anguish over evil high-frequency market makers and mean-spirited market makers....

;-)

 
kch:


I did not mean the accidental flow of money from one FI to another.

It is clear that there are reasons for asset flows (buying/selling) and they are not random, i.e. there are real people with real goals behind such movements.

Ahem. And my point is a bit different. There are strict obvious arbitrage restrictions on the degrees of freedom of quotes (three prices have two degrees of freedom).

As for people, they are statistically even more predictable than individually. In other words, the degrees of freedom of the masses of traders

are even less than arbitrage-linked prices :-)) That is why I would not put "real people with real aims" into "reasons". Especially

The fact that traders, regardless of their calibre, have only one or two goals, and not even two. To get a lot and that is the goal ..... ;)

1...131415161718192021222324252627...38
New comment