a trading strategy based on Elliott Wave Theory - page 36
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
Что же касается расчета показателя Херста, то мне совершенно непонятно Ваше нежелание высказаться по этому вопросу.
Так уже не раз говорил, что после идентификации выборки расчитываю показатель Херста для выбранного канала для того, чтобы сделать заключение о способе его (канала) участия в прогнозе.
Vladislav,
You must be kidding, repeating what you've already said more than once.
But if you're not and I was wrong, then I'll repeat for the third time the question around which the discussion is actually taking place.
solandr, when calculating the Hurst index, uses the scales of approximation errors and counts the spread as High - Low prices (and not errors). Is it correct?
Saving your time. Just yes or no ?
Good luck.
Vladislav 02.06.06 20:12
Yuri, you are absolutely irritated for nothing. I really did not understand your question. I am answering as you asked: for the problem you want to solve - that's exactly the estimate you need.
Good luck and good luck with trends.
Interesting point. And not understandable for me purely psychologically :) 5 minutes ago I did a test in analytical form which confirmed my suspicion :
1) Let have a random series Xi, where Xi = Close[i]-Close[i+1].
The variance for this series is Dx=X^2-X^2
2) A linear regression channel that approximates the series by the formula Yi=A*Xi+B, such that the line points on adjacent bars differ by delta_t.
Then we can write that Xi=Mi+ delta_t, i.e. we have a new random series Mi of the differences between Close[i] and the linear regression lines.
The variance of this series Dm=M^2-M^2
3) The transformation proves that Dx is identically equal to Dm.
This once again proves that the calculation of the Hurst index used in the Excel file coincides with the calculation that Vladislav uses. Only why didn't he say it at once, not to ask in circles? Either he was too lazy to explain, or he came up with it purely empirically...
Such tricks in our town :)
I downloaded the file, I saw the calculation scheme. Now I have to digest it.
The file is here - https://c.mql5.com/mql4/forum/2006/06/Hurst2.zip
The fact that the slope and the spread calculated with respect to the error (i.e. with respect to the linear regression channel) lead to H<0.5 is very understandable.
As I wrote above, the transformation y=a*x+b means a rotation of the coordinate system in which the regression line becomes the new Oh axis. There is no trend in this new coordinate system. Only market fluctuations around the time axis remain. And since the market is not subject to the normal distribution, then after removal of the trend component, only an antipersistent component remains, which corresponds to H<0.5.
By the way, it follows from the above that only 2 out of 4 calculation versions given by Rosh make sense: blue (where H is calculated in the usual way) and blue (where H is calculated relative to LR). The green and brown ones, where the sko is calculated relative to LR, and the spread is normal, or vice versa, do not make sense.
I wonder how they can be used to draw conclusions about market behaviour?
The first one. And here is more from the recent past, I specifically decided to check on the history in the place where there is a kind of "reversal" of the trend
Well, the probability of a reversal has paid off. It's moving up more than 150 pips.
There's just one thing I can't figure out. On the last chart Hurst=0.5127 and there is no trend.
Does it mean that in the absence of a trend the market always tends to the normal distribution
or is it only for this interval?