a trading strategy based on Elliott Wave Theory - page 187
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
With your kind permission I will ask you some more exact questions later (I am busy at work at the moment).
:о)
PS: If you do not mind, I can exchange e-mails. Mine is grasn@rambler.ru
I guess I have to get serious about DSP.
By the way grasn, remember our discussion about volatility? Neutron, as you can see, states the same as I do: volatility is estimated by the value of standard deviation.
But with a small difference: the standard deviation is measured by the difference between the opening and closing price, and the volatility is measured between the highest and the lowest price of the bar.
Yurixx, what is DSP?
Just kidding!
Hi. I don't get it - how you can express the Hearst index from the volatility value. To my mind - it's impossible. That's what I don't get :)
Yurixx, don't worry, I only pick up what I can understand in unfamiliar places. DSP is not yet in my range of interests, although I was trained as a radiophysicist :) But I don't see the need to refresh my knowledge yet. By the way, all these autocorrelation chips are described by Peters...
I sent a message to your inbox.
I sent a message to your inbox.
(Rosh beat me to it. :). A generic name for a rather large area. Evaluation, a matter of subjectivity. I, for example, have not yet engaged in spectral estimation and am even more of an amateur in this field. :о)
Got the letter.
Надо, видно, серьезно взяться за ЦОС.
Neutron, в приведенной формуле s0=SQRT(|SUM{High[i+1+k]-low[i+k]}^2|/{k-1})
есть кое-что непонятное. Возможно проблема в том, что запись формул в текстовом формате не отображает всех тонкостей. Не могли бы Вы пояснить
1. зачем нужен модуль суммы квадратов разностей, если это и так положительная величина
2. почему {k-1} в знаменателе стоит за знаком суммы, если суммирование ведется по к
3. почему High и low относятся к соседним, а не к одному, барам
Кстати, grasn, помните нашу дискусию по поводу волатильности ? Neutron, как видите, утверждает то же, что и я: волатильность оценивается по величине стандартного отклонения.
Still with a slight difference: the standard deviation estimation takes the differences between the opening and closing prices, while the volatility estimation takes the differences between the highest and lowest price in the bar.
Yurixx, what is DSP?
Yes, I understand that difference. That's why I wrote "estimated" and not "equal".
By the way, pay attention to the quote. I've completed my post with questions to you, but we've already moved to another page and you might not have noticed it.
DSP is digital signal processing. The apparatus you are talking about is in general much wider than DSP, but in terms of forex this difference is insignificant. I've never dealt with this field, so I have only the most general ideas about it, not beyond the Fourier series analysis course. However, with grasn's light hand, I became interested in it and I'm already reading something, but so far only the most elementary.
Yurixx, I completely disagree. Since when did spectral analysis become broader than DSP? It's like mechanics is broader than physics.
1. Why do we need the modulus of the sum of squares of differences, if it is already a positive value
2. Why {k-1} in the denominator is behind the sum sign, if summing is done by
3. Why high and low refer to neighboring, not one, bars
Yurixx,
1. it's not a module, it's just a bracket;
2. instead of k you should read n - I hurried when writing:-(
3. of course to one bar...
Damn, I'm so attentive! This is correct:
s0=SQRT((SUM{High[i+k]-low[i+k]}^2)/{n-1})
Generally speaking, knowledge of wateratility is necessary when estimating possible profitability of a certain TS or possible risks. When we want to estimate Hearst ratio, it is more correct to use the expression for the standard deviation:
c=SQRT((SUM{Open[i+1+k]-Open[i+k]}^2)/{n-1})
The algorithm of finding the Hearst ratio (M) is as follows
1. find the standard deviation values in steps of 1 minute in the range of 1 min to 1000 (for example);
2. knowing c1 (value of the standard deviation on the minutes) and the value at the next step (c2) solve the equation:
c2=c1*(t2/t1)^M1 => M1=ln(c2/c1)/ln(t2/t1) where t2 is the timeframe of two minutes. Then by analogy:
M[i]=ln(c[i+1]/c[i])/ln(t[i+1]/t[i]).
Thus, the Hurst index is a variable for this symbol and depends on the timeframe we work with. Indeed, on small timeframes the dynamics of the currency instrument price has a rollback character (autocorrelation coefficient is negative and as a rule exceeds -0.2 in absolute value), as a consequence, the Hurst index <1/2. On long timeframes (t>60 min) the dynamics of the currency instrument price has a random character (the autocorrelation coefficient is negative and, as a rule, close to zero), as a consequence, the Hurst=1/2.
As for the algorithm of Hearst index calculation described by you, it seems to me that it is not evident at all that using Open looks more correct.
Also, t1 and t2 should be different t/f in the formula. If we are really talking about a sliding window, then t[i+1] and t[i] refer to the same t/f. Therefore t[i+1]/t[i]=1 and the denominator of the formula for M[i] is 0.