From theory to practice - page 1505
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
Market memory - yes this is the darkest point, I was not very satisfied with Peters description to apply it, and noise colour and spectra also float by themselves, in fact only the diurnal rhythm is stable, other harmonics noticeably float and worse, they stratify and dissipate... I don't know what to do about it... There is only one thing for sure: looking at the news calendar one can suppose the "Joker effect by Peters" - a memory shift and a sharp fall in autocorrelation... Interesting options were suggested here on the forum too, lots of options, there's not enough time to try everything... From a purely visual point of view the movement should be no less than 200 bars of the working timeframe with a glance at the point of the last extremum and the scale of movements on the left in the history, but it is quite subjective, I understand how it sounds... so far i'm used to being guided by the chart form itself and the calendar...
can you list all these options?
maybe someone would like to try them.
I am not a mathematician, but only a user of mathematics, but it seems very strange to me! Maybe the problem lies in the definition of "earn"? For example, does it mean "earn on a certain interval"? Then you can stop the process when it is above zero... However, if the processes are added up in series, this will still have no effect because two processes stitched together is like one inseparable... another assumption is that it has something to do with incremental discreteness and rounding...
***
I think science has picked an unfortunate name for it.
Buddy, I'm just repeating my post:
If I cannot find the key, don't expect any more practice from me, it is enough for me to be ashamed of myself. And I do not recommend anyone who seeks it to enter the market without it.
Mate, I'll just repeat my post:
I will think about it...
can you list all these options?
maybe someone would like to try them.
First of all Maxim Dmitrievsky's article, I don't remember other links, it was scattered on branches, generally through autocorrelation and spectra, there was also an idea about using wavelets, I can't find the link to the post too, but instead here is a link to the description (see 2.3 The wavelet methodology)https://www.cairn.info/revue-finance-2014-2-page-57.htm and here are works of Chinese mathematicians: https://www.researchgate.net/publication/327192087_Memories_of_the_Gold_Foreign_Exchange_Market_Based_on_a_Moving_V_-Statistic_and_Wavelet-Based_Multiresolution_Analysis with wavelets it's good that wavelets don't care about non-stationarity and nature of data but unfortunately there mathematics is so hard to beat ((
Of course I am not a mathematician, just a user of mathematics, but I find this very strange! Maybe the catch is in the definition of "earning"?
The catch is that the author of the thread is confused about the terms.
In this case, he is referring to a stationary, incremental type process. Not the integrated one that we are actually trading.
You can also work with the price like this
The problem is that the author of the thread is confused about the terms.
In this case, he is referring to a stationary, incremental type process. Not the integrated one that we are actually trading.
Well yes, a stationary process is traded, and the theory is constructed for a process with stationary increments.
"Gravity of modular vortices".
The problem is that the author of the thread is confused about the terms.
In this case, he is referring to a stationary, incremental type process. And not integrated one which we actually trade.
Well, then it turns out to be some kind of pipsing (by time)...
multiplicator:
such a process is also called random
***
I think science has an unfortunate name for it.
Of course the processes may have different distribution parameters... but I was thinking... Maybe the problem is that the process is of this type (see pic) with a very strong kurtosis and very large tails... then it's not hard to imagine a scalping strategy like breaking the vertical dotted lines... Technically it is cheating anyway and not true because a breakthrough stop order will be executed inside of one increment (one tick) and that's not allowed....... - you need at least two ticks to place and execute an order and it means the second tick will be counterclockwise with the first one and it already violates the independence of observations (i.e. increments are not independent) ...... in reality ticks may well be independent, but according to the problem we need independent - it turns out hells - one conflicts with another (if I am not mistaken)
However some versions of TSC, as far as I remember, allow some dependence, but there are all sorts of additional conditions (that if a little bit dependent - you can).
In short, the riddle of the century - how to make money on randomness? - or get a nobel prize! or go to the factory hehe
You can also work with the price in this way
This is the multicone with a high attenuation factor
I understand there is a desire to move away from the trend...... Maybe it makes sense to consider not the price chart, but for example the signals chart.....
Given: A set of nonstationary series, the distribution parameters are floating, there is no cyclicity
The task: to obtain a trend-stationary nearly monotonically growing series (equity) from the original series
Admissible operations: cutting of fragments from initial series (position opening-closing), multiplication of a fragment by any number, summation
Peeping into the future is of course forbidden.
Who solved the problem will go to Malta