Put in a good word about the occasional wanderer... - page 25

New comment
 
alsu:
I meant the influence of principle and methodology, not from the point of view of practical achievability of the result. The algorithm works if it is possible to estimate probabilities, and how it works - whether it allows to trade profitably or not - is a subject of research, I cannot answer this question in advance.
If you remember again the original problem, the conditions are different there. But it is possible to estimate the probability, and if you do it, it will be less than 50%.
 
alexeymosc:
If you think back to the original problem, the conditions are different. And the probability can be estimated, and if you do it, it will be less than 50%.

The problem is, of course, modified, but the solution principle will be similar.

Probability in the original problem, if you remember, strongly depends on m. So there is some hope that in the modified problem with reasonable m it will be possible to get the cherished 71%.

 
alsu:

The problem, of course, is modified, but the solution principle will be similar.

Probability in the original problem, if you remember, strongly depends on m. So there is some hope that the modified problem with a reasonable m will also be able to get the cherished 71%.

If we remember the Paradox from the "problem of ruin", the law of arcinus and possible causes of "fat tails" in increments - profitability may take place.

It was not for nothing that hrenFX used pending orders and was interested in mathematical models of such strategies (seepage 19) ...

;)

 
avatara:

If you remember the Paradox from the "ruin problem", the law of arcinus and the possible causes of "fat tails" in increments - profitability may have a place.

and I have already learned how to cut off thick tails... I analyze them separately with the main series, you actually get two series with completely different characteristics: one is a Gaussian wandering, the other is a generalized Poisson flow.
 
alsu:

The problem, of course, is modified, but the principle of the solution will be similar.

The probability in the original problem, if you remember, strongly depends on m. So there is some hope that even in the modified problem with a reasonable m it will be possible to get the coveted 71%.

Well, one must think, and most importantly, count on empirical data. You see the point: the problem of the princess is solved analytically due to the assumption of a priori data independence (here also stationarity of the series) plus some restrictions in order to complicate the solution. And Forex series is a special subspecies of the family of financial series, belonging to the realm with such kind of probability density function... On the one hand one cannot disagree, however, on the other hand one cannot disagree (C Strugatsky)... You cannot solve analytically, and the initial measurements showed that waiting for the price to return to the right level, in its pure form, is not reasonable, if only because it is over sitting of drawdowns, leading to the fact that in some cases (OK, they are more than 50%) the price will reach profitable level, but in other cases it will drawdown the depo so much, that MO will equal, oh miracle, - minus spread. I modelled this mess in excel on the clock, with a wait time of up to 10 hours. Exactly minus the spread, my gentlemen. (Full stop.)
 
A webinar on probabilities. Fourier and Hurst are also mentioned.
 

nice feature to yourself...

Это простейшее дифференциальное уравнение, имеющее точку, в которой вид решения меняется с колеблющегося на экспоненциальный.

:)

 

B. Berezovsky is a head...

;)

 

Harmonious theme...

;)

 
Frankly surprised, I don't know what kind of prince you're looking for in a princess. And why are you trying to give the princess a rake so she'll turn over a third of the pile of crap looking for a pattern. That's not how the market works. Princes don't just show up at random, but under very specific conditions. If you see those conditions, grab the first prince that comes along and you'll be happy. ;)
1...181920212223242526
New comment