a trading strategy based on Elliott Wave Theory - page 198
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 point is that when I calculated Y[i]=Open[i] for series, I removed not only the constant component, but also the trend component. That is, I calculated a linear regression for the entire interval and subtracted y=a*t+b from each term of series Y[t]=Open[t]. In theory, I should have obtained a sign-variable numerical series of oscillations around the regression line. Strange, I'll have to check what it actually turns out to be.
P=1-2^(N-1)*P{1-p[i]}
Interesting formula. Perhaps for dependent indicators we should get something like this:
unit turns into a matrix of correlation coefficients,
p[i] turns into a vector and product turns into determinant.
Sergiy, what can you say about the correlogram of indicator with price series. I had an assumption that this correlogram should reflect the usability of the indicator in the market. If r[k]=0 for all k, then it's obvious that the indicator doesn't correlate with the price and its use is equivalent to flipping a coin. If it is greater than 0, it seems that something can be extracted from it. Ideally (Vesnukhin's fantasy :-) the correlogram maximum should have been in the future, then one could say that this indicator has prognostic properties. But the result I got (constant positive value) is somehow difficult to interpret.
Maybe I overestimate the correlogram as a source of information or evaluation ? Or I am doing something wrong.
By the way, why is the formula for calculating the autocorrelation coefficient you gave different from the one you gave before ? I don't understand this new formula at all. And what does it mean
i?
Explain: by the first case we mean the series X[i]=Open[i] and by the second case X[i]=Open[i]-Open[i-1]? Or do you mean something else ?
How did you remove the constant component from the first series ? By subtracting the mean over the whole interval ? Or by moving average ? Or another way ?
PS. Happy New Year and good luck in the new year.
I'm a little sobered up from a string of carpal tunnels from the 26th. Ugh!
There was a mistake in the previous message. The correct way is:
1. The correlation function between two time series X[i] and Y[i] is defined as:
F=SUM{X[i]*Y[i]}/SUM{X[i]*X[i]}, where summing is performed over all the members of i=1...n.
2. I get FFT on the required TimeFrame (TF) only from minutes by the scheme:
X(TF)=Open[i*TF]-Open[(i-1)*TF]
Y(TF)=Open[(i+1)*TF]-Open[i*TF]
The obtained series F(TF) decreases monotonically and shows the dependence of the expected price jump on the previous one as a function of the TimeFrame.
3. The correlogram is plotted on the current TimeFrame by the following scheme:
X=Open[i]-Open[(i-1)]
Y(j)=Open[i+1+j]-Open[i+j]
The obtained series F(j) is sign-variable and shows the dependence of the expected price jump on the arbitrary previous j on the selected TimeFrame.
I don't think I got it wrong...
One could estimate the amount of alcohol (in gecalitres) drunk at these parties, based on the correlation of this amount with the number of discrepancies between this post and the previous one.
However, hard ... :-)))
No more questions as corporate parties are followed by private ones.
Happy New Year !
PS: Sergey, I've been trying to check the calculation of the correlation (SUM{x(i)*x(i+k)}/SUM{x(i)^2}). The formula I use to calculate it is almost the same, but with a slight difference:
I use this formula for the correlation function of two time series X[i] and Y[i]
F=SUM{X[i]*Y[i]}/{|X|*|Y|},
where |X|=SQRT{SUM{X[i]*X[i]} and |Y|=SQRT{SUM{Y[i]*Y[i]},
and the factor 1/n is omitted, as it is in both numerator and denominator.
Why would that be the case? This formula is used just for any signal. There is no such limit, where did you get it from?
Where to get them?
We have only one indicator: Price.
Why do we need indicators?
A year and a half ago I carefully studied Nyman's encyclopaedia and came to a conclusion that all known indicators are very primitive and have a very similar method of calculation.
Increasing their quantity increases the entropy of results and the equity graph becomes as unpredictable for the TS author as the indicator value graph. :D
Where to get them? we have only one indicator:
Price. Why do we need indicators? a year and a half ago, I thoroughly studied Nyman's encyclopedia and came to the conclusion that all known indicators are very primitive and have very similar calculation methods. their quantity increases the entropy of results, and the equity graph becomes as unpredictable for the TS author as the indicator values. :D
1. Why would you use several indicators at the same time? THIS IS DIRECT!!!
Answer the question: "If you wrote an indicator that makes a stable
profit, would you put it out there for everyone to "use"?" :))) Mda-a-a-......
see... :)))
It's easier to check in practice all available indicators and create your own "reliably working" indicator on the basis of this one
.
Many people don't want to spend time on checking, they read branches on different forums
(like this one) and try to find an answer to this question...
"How to make money"? BUT....................................................................
2. About this:
2 Rosh
What difference does it make to me to buy 10 pairs at 0.1 lot or 1 pair at 1 lot, if p=0.55 for all instruments?
Just for the sake of diversification, to reduce possible drawdown? It makes no sense.
The slope angle of the profit growth curve will not change, and therefore the rate of equity growth will not change either. But p=0.8 will give a completely different picture.
Moreover, I doubt that any indicator will have the same p-value at all pairs. The nature of price movements is different, and I suppose p will also be different. In this situation it is better to choose the pair where p is maximal rather than diversify.
I checked both variants in practice, I consider more acceptable and grounded Rosh's position
Yurixx no offence :)))
Sincerely,
Alexey