Predicting the future with Fourier transforms - page 49

 
2 USSR, Svinotavr: the flooding has been taken down. That's not all, I'm just getting started.
 
Reshetov: I am only talking about the advantage. If the amplitude of the first harmonic exceeds several times the spread, the advantage will necessarily be reflected in the expectation. Otherwise, there is no fish here, i.e. if the amplitude is on the edge of the foul.

Hmm... "the amplitude is several times the spread"? How is that related?
 
LeoV:

Hmmm... "The amplitude is several times the spread"? How is that related?

Well, the amplitude is in pips and the phase is in time.

It's trivial, the amplitude is always calculated in the units in which the signal is measured.

 
Reshetov: Well, the amplitude is in pips and the phase is in time.

It's a matter of principle, the amplitude is always calculated in the units in which the signal is measured.

No, that's not the point. Why, if the harmonic amplitude is several times greater than the spread on historical data, this harmonic will be profitable in the future? What does this have to do with the expected payoff?
 
LeoV:
No, that is not the point. Why, if the harmonic amplitude is several times greater than the spread on the historical data, then this harmonic will be profitable in the future? What does it have to do with the expected payoff?

Well, if the amplitude is currently at an extremum, even a drunken hedgehog knows that it can only move towards an extremum with the opposite sign. Its future, i.e. the direction of movement is already known before the next extremum. It's trivial.

If the amplitude is equal to several spreads, then it will give an advantage of at least the size of the amplitude, which is also trivial.

 
USSR, chill out with the heavy and endless flubbing for 24 hours. It's a long way to go.
 
Reshetov:

Well, if the amplitude is currently at an extremum, even a drunken hedgehog knows that it can only move towards an extremum with the opposite sign. Its future, i.e. the direction of movement is already known before the next extremum. It's trivial.

If the amplitude is equal to several spreads, it will give an advantage of at least the size of the amplitude, which is also trivial.

The "tricky thing" about the market is that the amplitude of the first harmonic estimated from past observations may be even smaller than the observed one... or the sum of a couple of other harmonics will throw the price much further.

;)

 
Rorschach:

...cycle frequency in amplitude...

So it could well apply to the market.

How's that?
 
Reshetov:

Well, if the amplitude is currently at an extremum, even a drunken hedgehog knows that it can only move towards an extremum with the opposite sign. Its future, i.e. the direction of movement is already known before the next extremum. This is trivial.

If the amplitude is equal to several spreads, it will give an advantage of at least the size of the amplitude, which is also trivial.

Yuri, I'm sorry, but apparently a drunken hedgehog understands more than I do. I'm not even asking about what "amplitude is at an extremum" means.
 
tara: I'm not even asking about what "amplitude is at an extremum" means.
Well, you've got it all figured out...