_Market description - page 30
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Yes, I agree, but I hope you won't deny that if there is a periodic function in this region it will show up in the spectrum.
Prival, what is the "spectrum" here in this sentence? What you get after Fourier is not the true spectrum of the function, but just an approximate interpolation by a set of multiples of harmonics. And then, where in real life, let alone in trade, have you seen a periodic function? Even a crystal oscillator, the finest component of radio electronics, its harmonics ARE NOT CRATIC (not 2.00000 and 3.00000, but 2.0000032456, 3.00000459231 etc.), so it is not periodic (not to mention other factors such as noise, which all together lead to such a phenomenon as jitter, ie frequency oscillation of crystal oscillator).
Personally, I am very glad that you have started to agree here with the strong LIMITATION and particularity of the Fourier method and so Kotelnikov's method.
Insanely glad. Honestly. ((C) Belmondo).
I apologize that I have entered, probably not timely, but I could not keep from it as I have the reflections on the given question on which now actually I work, besides the raised theme so-to-say "is close in spirit". it is in general a preamble...
and actually "ambula"... the cochlea in a human ear (and other mammals) "as though" is capable to carry out function of the frequency filter, and in fact if to generalize "it seems that" represents usual spectrum analyzer. further I think few will find objections to that practically any person (if it not deaf certainly) is capable even in the conditions of raised noise (on production for example) to allocate even very difficult BUT PERMANENT signal, (voice of the partner for example) which level is obviously LESS than level of this industrial noise. it is first ...
second, if you run a quote stream through an amplifier in "accelerated mode" you can clearly hear intermittent components that are even more numerous than the noise.
Summarizing all abovementioned it seems reasonable to use Fourier transform for splitting the kotir into a spectrum, with the following giving of this spectrum to the NS input and making the decision to buy or sell.
PS. to plagiarize nature's idea for its own purposes. :)
Nobody argues that Fourier transform can describe with 100% accuracy any signal: periodic, non-periodic, random or deterministic. Here the straight a+b*x was mentioned. A straight line will also describe it. But just because you have reproduced the signal with 100% accuracy does not mean that you now know how the signal will behave in the future. This can be proven very easily. I will give you some random numbers out of my head. And ask you to predict their future values. You apply the Fourier transform, decompose them into sines and cosines, verify with 100% accuracy this trigonometric model and then extrapolate it. Do you think your predictions will coincide with the next series of numbers I give you from my head? If you're sure it will, then open a fortune-telling business.
So, uh... Are you interested in this test? Here are the first 10 numbers:
12.3 15.6 2.7 3.9 21.0 14.3 17.8 11.0 9.7 19.0
I'll let you apply the neural network if you like.
You shouldn't take the numbers out of your head. Better to generate them somehow. It has been proven that humans cannot come up with a random sequence in any way.
The better the chances are for Fourier fans to predict the sequence. =)
gpwr писал(а) >>
Do you think your predictions will coincide with the next series of numbers I give you from my head...?
I think that "sequences of numbers" repeat in the market from time to time, for example when a big player enters the market it is always noticeable... and if some "sequence of numbers" appears chaotically in the market for half a month now, I believe that Fourier coupled with NSK would be a great tool for catching it. and also for finding such "sequences of numbers".
Try to think of something distracting (say, the last corporate party :) ) without looking at the paper, write down a long "sequence of numbers", say, two hundred.Or if you're good at five-fingered typing, you can also type on the keyboard.
Similarly, members of this forum (and not only this one) argue about trend and flat, assuming these concepts are defined. But they only see the tail of the elephant. One should look at the elephant as a whole! Look at what the "primitivist" Figar0 says:
Yes, sic! Both are unlikely, but very likely one of them (if it's a true trend). Now try to guess where I'm going with this? To a trend with a flat! And there's the key concept (and there's Semenych's articles on cluster inducators, in which this idea is central):
A trend only happens on a currency, not on a pair.
A flat should probably be on both currencies of the pair, but again not on the pair.
Let's go like this ... if we're talking about the market model ...
And what do we actually know... about the market...
Let's just agree from the beginning...
1) we are looking at 8 currencies, i.e.
EUR GBP USD CHF JPY AUD NZD CAD and everything related...
Other instruments will be left alone
2) we know the formula for calculating the cross-rate (e.g. GBP/CHF=GBP/USD*USD/CHF)
3) We obtain 28 pairs accordingly...
4) The spread is also not considered for the accuracy of calculation ...
5) to 4) that is, we calculate everything from the dollar rates
(spread is the smallest in all brokerage houses ... let's calculate it accurately ...)
Nothing new so far...
6) any market movement is always 100% correlated with 2 currencies
(that is, if we consider a time moment T = 0 and T = 1, there is always a pair in which one of the currencies was falling / rising relative to all others ...)
7) If it is so ... 6) ... then we can always find a Pair, which in the average (in different directions) is moving faster than all the others ...
Accordingly, it and trade ...
BUT IN WHAT AND HOW DO WE COUNT...?
20099 писал(а) >>
Here's what "primitivist" Figar0 says:
...
6) every market movement is always 100% correlated to 2 currencies(i.e. if we consider a time moment T=0 and T=1, we will always find a Pair in which one of the currencies was falling / rising relative to all others ...)
7) If it is so ... 6) ... then we can always find a Pair that moves faster than all the others in the average (in a different direction) ...
Accordingly it is what we trade on...
1. Correlated not relative to two currencies, but to two pairs of currencies with a positive non-linear correlation (you can't make any money on linear correlations). If you look at points in time, one of the pairs is rising faster than the other, and the other is falling faster
2. Do not even try to identify the pair - favorite, which moves faster than others in any direction. It may turn sharply against the previous trend and switch to the opposite trend at any time.
For a smart hedge, it is necessary and sufficient to place a long position on the pair that is growing faster, and a short position on the one that is falling faster. When the price goes up, we will earn on the fast-growing pair and lose money on the slow-growing one. When the price goes down, we will earn on the fast-falling (slow-growing) pair and slowly lose money on the slow-falling (fast-growing) pair. That is, no matter where the price goes, the final profit is statistically justified. Of course due to hedging (insurance) we will lose a part of profit, and quite a big one at each movement compared to what we could potentially gain. But in order to take profit without hedging we would need to predict every movement, which is either unrealistic or very unstable.
Based on the above principles I created a multicurrency TS, the results of which you can see in the thread:
Project - Portfolio Manager
Calculations are made using hedging pairs:
EURUSD + GBPUSD
EURUSD + AUDUSD
EURUSD + NZDUSD
Since all pairs have a linearly proportional pip value to USD, statistics should be calculated in pips.
1. Correlated not relative to two currencies, but two pairs of currencies with positive non-linear correlation (you can't make any money on linear correlations). If you consider points in time, one of the pairs is rising faster than the other and the other is falling faster
Agreed... )))) in a closed system it CANNOT be different :))) (sorry, I am not a mathematician...) BUT the point does not change...
2. Do not even try to identify a pair - the favorite, which moves faster than others in any direction. It can easily reverse at any moment, against the past trend and move in the opposite direction.
I disagree = you can calculate= anything...
sorry to repeat what "THIS" is the most =NO we should calculate...??(it's about TREND... FLET... IMPULSION... I'm for the latter...)
especially... I wrote that in (average = all indicators are based on that principle) ... only in WHAT??? (what do we take as a unit???)
For a smart hedge, it is necessary and sufficient to place a long position on the pair that is rising faster, and a short position on the one that is falling faster. When the price goes up, we will earn on the fast-growing pair and lose money on the slow-growing one. When the price goes down, we will earn on the fast-falling (slow-growing) pair and slowly lose money on the slow-falling (fast-growing) pair. That is, no matter where the price goes, the final profit is statistically justified. Of course due to hedging (insurance) we will lose a part of profit, and quite a big one at each movement comparing to what we could potentially gain. But in order to take profit without hedging we would need to predict every movement, which is either unrealistic or very unstable.
Based on the above principles I created a multicurrency TS, the results of which you can see in the thread:
Project - Portfolio Manager
Calculations are done on hedging pairs:
EURUSD + GBPUSD
EURUSD + AUDUSD
EURUSD + NZDUSD
Since the pip value of all pairs is linearly proportional to USD, we should calculate statistics in pips.
>> I'll see...
I created a multi-currency TS on the above principles, the results of which you can see in this thread:
Project - Portfolio Manager
looked at ... (sorry I didn't understand the point of this thread...)
continue (discussing "MARKET MODEL") on the spot...
looked ... (sorry I didn't understand a fucking thing in this thread...)
It's clear that some people don't get it. Then I won't bother you any more.
Good luck finding a description of the market and its models!