Market phenomena - page 50

 
faa1947:

You're the smart one.

I am not questioning your mental capacity and qualifications, but your model is wrong from the very beginning, in its very formulation.

Yeah, sorry about the name-calling, but I'm kind of finishing the last lines, like in a ban. Soon they'll realise I'm farnsworth, and I'll get banned again. You'll be talking to hexogen by now.

 
OK, here's a phenomenon for the sake of refreshment - if we plot the volatility distribution by the minutes of the hour (High-Low), the peaks are at 0, 15,30,45,59. And even more - in those minutes there is a stable movement in the direction of the trend, but about the average spread. Moreover, the trends are formed with a period of up to one hour.
 
_Farnsworth:
That's not what mine are called, honestly.

If the variance is stable, where does the thick tail come from?

Outliers of variance make low-probability events probable. Fractalists seem to have proved that.

 
Svinozavr:

(shrugging) Start.

I don't get it, am I first, second or ? ))) // Serge. Relax, will you! Everything is relatively (how else could it be!) normal. I, for example, am interested in reading you. I'm not lying.

I haven't gotten to you yet... I feel a moderator's going to stop me. Anyway, good with a shotgun beats evil...
 
faa1947:

If the variance is stable, where does the thick tail come from?

Outliers of variance make low-probability events probable. This seems to have been proven by fractalists.


It's not really about instabilities, it's about variance/volume dependencies on previous values of the variance/volume. And if the vola varied but without certain dependencies, it would still come down to HP
 
_Farnsworth:
I haven't gotten to you yet... I feel a moderator will stop me. In general, good with a shotgun will defeat evil...

))) Good should come with fists. At the very least (what, what end?)) with a shotgun.

===
It's a shame I haven't made it.

 
... ...come along, break out a birch tree and have fun...
 
Avals:

It's not really about instability, it's about variance/volume dependencies on previous values of variance/volume. And if the vola varied but without certain dependencies, it would still come down to HP.
I'm not quite sure what dependency we're talking about. Usually one starts ARCH modelling after removing dependencies from the series. I'm not aware of any other methodology.
 
faa1947:
I'm not quite sure which dependency you're talking about. Usually you start ARCH modelling after removing dependencies from a row. I am not aware of any other methodology.

Actually, ARCH modelling is the modelling of a wave based on certain dependencies on its historical values
 
faa1947:

If the variance is stable, where does the thick tail come from?

Outliers of variance make unlikely events probable. Fractalists seem to have proved that.

Did I write about stable dispersion? Where? By the way, can you prove that you have it at all?

faa, my favourite nickname got banned the hell out of me. I won't be under the creepy _farnsworth for a long time :))) I'll kick svina a couple more times and then go mind my own business. It's important to emphasize - mine. I just realised that trying to give my brain a recognisable look isn't really my thing. And it's a hassle.