From theory to practice - page 670
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
The conclusion is very simple and adequate - 90-95% of all trends or as I like to put it "directional movements" originate in a state of complete uncertainty and tend towards complete chaos.
That's right. That's how the market works - from chaos to self-organisation and back again. Calculating the moment of transition is incredibly difficult, but it is possible.
I remain with my deepest conviction that the magic key is non-entropy/entropy. I just can't figure it out :)) I'm getting incredibly stupid here. I'm in the firm embrace of the local underdogs and retards...
For some reason I am running from one idea to another. For example, the flat channel (it may be as wide as I want) with the multiplier 1.6. The signal to buy on EUR/USD pair is 17:56 GMT+3. May be it is vice versa a signal to place a pending sell order from the current price + N known points. Since the price has reversed at that level. But this is just an assumption.
And how can this be identified? Alexander, you are an expert in distributions, and you demand that any suggestions be justified physically and mathematically. How can both traditional and non-parametric statistics of the DISTRIBUTION of a random variable be related to "memory"? After all, no property of a random variable's distribution changes if you rearrange the order in which its values appear as you wish.
For example, in this way which is universal for the rate increments: we order all the available increments in the sample by their values from the lower to the upper ones. The corresponding rate will first decrease, then increase. Irrespective of the multimodality of the distribution. What will happen to the "memory" in this case - complete chaos, it will disappear, if there was one. Although neither asymmetry, nor kurtosis, nor mean and variance will change at all.
It remains to rely on some property that is not in the distribution. For example, time - and then one can try to think about the growth of entropy, although in case of currency interventions it will obviously decrease. It is not enough to analyse only distributions, you need a link to the LETTERS. You think asymmetry and kurtosis are related, tell me on what basis?
So that your decisions can be justified physically and mathematically.
You use the name "diffusion coefficient", while it did not arise by chance - processes of diffusion, heat conduction and filtration are described by the same equations of parabolic type and in absence of perturbations the transfer potentials in them are dissipated in space with time. At the same time, entropy increases. By the way, the square root law also works, remember the similarity of thermal nonstationary processes by Fourier criterion at/x^2. Tossing a coin can also be described by the heat conduction equation. And what is the basis of your reliance on kurtosis and asymmetry?
You will laugh, but my trading is based precisely on time. Price is used indirectly. By the way, I'm not an expert on distributions and generally - Alexei.
For some reason, I'm going from one idea to another.
:))) And I've already calmed down.
Let the TS work in a channel relative to the median in the window=24 hours and screw it.
I have not managed to find a magic key to filter out losing trades.
I will hope for the fate and a miracle - if some flying genius like podotr or CheGevara will say: "Okay, old man - look at page 666 of this book, and everything will become clear to you".
In the meantime, I wash my hands of you and kiss your ass. I didn't make it. I repent.
:))) And I've already settled down.
Let the TS play in the channel relative to the median in the window=24 hours and screw it.
I have not managed to find the magic key that filters out unprofitable trades.
I will hope for the fate and a miracle - if some flying genius like podotr or CheGevara will say: "Okay, old man - look at page 666 of this book, and everything will become clear to you".
In the meantime, I wash my hands of you and kiss your ass. I didn't make it. I repent.
Alexander, you are not you here, but an apologist for an idea. Leaving the thread without specifying the prospects of the idea, you leave the thread in shame. Shall we try to do it with honour?
:))) And I've already settled down.
Let the TS play in the channel relative to the median in the window=24 hours and screw it.
I have not managed to find the magic key that filters out unprofitable trades.
I will hope for the fate and a miracle - if some flying genius like podotr or CheGevara will say: "Okay, old man - look at page 666 of this book, and everything will become clear to you".
In the meantime, I wash my hands of you and kiss your ass. I didn't make it. I repent.
When was the bad deal on the euro in September?
Look in this file how i.EUR was against i.USD at that time
Alexander, you are not you here, but an apologist for an idea. Leaving the thread without giving a perspective on the idea is a disgraceful abandonment of the thread. Shall we try to do it with honour?
No, well, there's no shame - there's a solid +0% profit.
I.e. if you just use the proportion of price~root of time, then there can't be any drain in principle. It's proven in this thread. So is happiness in the form of money :)))
I will read the works of old Gunn at my leisure. He also took this proportion: price ~ the root of time, and used some circles and squares as the basis for it. And he found the magic key of profit, unlike me :)). Maybe I'll find something interesting there. I hope so.
And when was there a bad deal on the euro in September?
Look here in this file how i.EUR was against i.USD at that time
Nah, that's it Zhenya - I'm reading Gunn and don't want to get distracted. Sorry.
No, that's it, Gianni - I'm reading Gunn and I don't want to be distracted. Sorry.
I could be wrong, at least what I have seen in the calculations is that the sum of price increments over N bars equals the same difference between the two prices of that period.
That is the increments for 1440 minutes = Price 0 - Price 1440.
Well, maybe I got such a plot, but it is more likely to be so. It follows - but why count the sum of increments if it is faster than the difference.