From theory to practice - page 1323
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
Well, the entropy in the moving window is 150 (black) and the same but for SB (red), for returnees
it is clear that a market retour differs from a standard SB and the entropy decreases on the trend segments, and approaches the pure SB on the flat segments
It happens due to tails in returns but not cycles. It is especially visible at the orange area where entropy reacts to sharp spikes (but not to cycles). Can this be called "memory"? Of course not.
If you add tails randomly to the SB, it will be almost the same. I'll check that later (I don't know why).
The other thing is that the volatility is clustered (outliers appear with a certain frequency), so the entropy changes smoothly. Well that's heteroscedasticity as usual.
Well, the entropy in the moving window is 150 (black) and the same but for SB (red), for returnees
it is clear that a market retour differs from a standard SB and the entropy decreases on the trend segments, and approaches the pure SB on the flat segments
It happens due to tails in returns but not cycles. It is especially visible at the orange area where entropy reacts to sharp spikes (but not to cycles). Can this be called "memory"? Of course not.
If you add tails randomly to the SB, it will be almost the same. I'll check that later (I don't know why).
The other thing is that the volatility is clustered (outliers appear with a certain frequency), so the entropy changes smoothly. Well that's heteroscedasticity as usual.
Read my comments all the answers to all your questions are in my comments.
heteroscedasticity of returns exactly, i.e. emissions become larger, for example. Apparently directional movement is sometimes confirmed by volatility clustering, although not necessarily
Plus the serial correlation appears. In the sliding window it's all perceived as a deviation from the SBRead my comments all the answers to all your questions are in my comments.
Not heteroskedantic) for who told you that
directional price movement can't be random?)
no it can't
The price goes on and on, not deviating even slightly from its course, from point to point.
Well, the entropy in the moving window is 150 (black) and the same but for SB (red), for returnees
it is clear that a market retour differs from a standard SB and the entropy decreases on the trend segments, and approaches the pure SB on the flat segments
It happens due to tails in returns but not cycles. It is especially visible at the orange area where entropy reacts to sharp spikes (but not to cycles). Can this be called "memory"? Of course not.
If you add tails randomly to the SB, it will be almost the same. I'll check that later (I don't know why).
The other thing is that the volatility is clustered (outliers appear with a certain frequency), so the entropy changes smoothly. Well that's heteroscedasticity as usual.
Well done, Maxim!
Window = 150 CLOSE M15? Did I get it right? I.e. = 2250 minutes? Is 1.14103 the average BP entropy value?
Of course I can't give you any assignments, but if you get a chance, please make a table of average entropy values for sliding windows in multiples of 1 hour:
60 minutes (an hour), 120,180, 240, ..., 1440 minutes (a day).
We need to find the window with the minimum average entropy value BP.
Well done, Maxim!
Window = 150 CLOSE M15? Did I get it right? I.e. = 2250 minutes? Is 1.14103 the average BP entropy value?
Of course I can't give you any assignments, but if you get a chance, please make a table of average entropy values for sliding windows in multiples of 1 hour:
60 minutes (an hour), 120,180, 240, ..., 1440 minutes (a day).
We need to find the window with the minimal average value of entropy BP.
got it, I will do it... I wonder where it can be used in TS
not quite understand - wikipedia uses natural logarithm in the formula, although it says that the bases can be different... well, left the natural logarithm for now
No, it can't.
the price continues to rise and fall, not a bit deviating from its course, from point to point
well, then explain to me why I can simulate such a thing with a random number generator?)
There's a lot to say) but what I want to say is this:
I may be wrong, but numbers never lie)
i believe in numbers, dry statistics and nothing more and i advise all who read me to do the same)
I understand, I will do it... but I wonder where it can be put into TC.
I don't quite understand - wikipedia uses the natural logarithm in the formula, although it says that the bases can be different... well, I left the natural one for now
Let it be as it is. And it's so clear.
If such a sliding window is found and has a stable minimum mean entropy for various currency pairs, then it is the window that should be used in your TS. It is in this window that the maximum difference from the SB will be observed.
I have found such a time window (though by different methods) and I am interested to compare it with independent studies.
Well, then explain to me why I can use a random number generator to simulate this?)
but it doesn't look like it yet - the spread is narrow.
Can you widen it to, say, 70 pips and find some kind of pattern in this movement?
but it doesn't look like it yet - the spread is narrow.
If you want to expand, for instance, to 70 pips and find some kind of pattern in that movement.
How many charts do you need?
I can draw twenty or a hundred))
but the point won't change... not at all
What does that have to do with the spread? For example, you may analyze one-hour charts.)
We're talking about the fact that the direction of price movements is not random at all)
I'm speaking on behalf of statistics and provide clear examples.