From theory to practice - page 1101

 

There are two risks in the market. Informational and monetary. Take your pick. And if one place is itching to count the time between ticks, their distribution or whatever........ Scientists in the world with heads as big as the whole forum beats and shish....

Money risk will lead to direct losses. Informational - to the unexpected.))

 
secret:

The simplest case is the dependence of the increment with time t+1 on the increment with time t.

Time is involved here in the most direct way.

But the time between quotes is non-linear, not +1! Even when you read equal intervals, the time inside them is exponential!

Anyway, I'm sensing some kind of mystery here...

In the meantime, today's first trade on GBPJPY was in the plus. Let's live!

 
Alexander_K:

But the time between quotes is non-linear, not +1! Even when you read in equal intervals, the time within them is exponential!

Anyway, I'm sensing some kind of mystery here...

In the meantime, today's first trade on GBPJPY was in the plus. Let's live!

I mean, I don't see it in the red.

;)

 
Alexander_K:

But the time between quotes is non-linear, not +1! Even when you read in equal intervals, the time within them is exponential!

Anyway, I'm sensing some kind of mystery here...

In the meantime, today's first trade on GBPJPY was in the plus. Let's live!

Of course, time can be as non-linear as you like, it's market time, not astronomical time.

OK, in general terms let's rephrase it like this:

"the dependence of the increment with number n+1 on the increment with number n."

The translation of t to n can be varied, and the "increments" are not necessarily ticks. It can be for example n+1 - increment for the next session, and n - increment for the previous one.

 
The second GBPJPY trade didn't seem to go through)))
 
Alexander_K:

But the time between quotes is non-linear, not +1! Even when you read in equal intervals, the time within them is exponential!

Anyway, I smell a mystery here...

In the meantime, today's first trade on GBPJPY was in the plus. Let's live!

It's not time - it's distance. It's not in pips or pips.

 
Maxim Kuznetsov:

not time - distance. It's not in points or pips.

I don't know... The thread has long since become a collection of vague phrases...

I write concrete things, only no one gets it, being weakened by the struggle with the market.

Specific things are research drawings and graphs. R&D, so to speak...

As an example:

This is a histogram of the time intervals between the CLOSE prices of the previous and OPEN prices of the current non-empty (when there was at least one incoming tick) S2 bars.


I don't know why they suddenly "cut out" times - 7, 14, 21, ... Probably, I was mistaken somewhere, but the essence does not change.

We see the obvious Pascal distribution (or Erlang for the discrete case). And such a picture - for any TF and for tick quotes!!!

Now it becomes clear why the classical methods of working with signals do not work in the market (calculation of autocorrelation coefficients, Fourier series, etc.). Yes, simply the mathematical apparatus of these methods is designed for uniform discretization of raw data.

I can't help thinking that we should switch to another coordinate system, another event space (like Minkowski space) where time would be taken into account and the Grail would be in our pockets.

And now I'm sorry - I have to get ready for work, i.e. work. To earn a penny for our daily bread...

 
Alexander_K:

I don't know... The thread has long since become a collection of vague phrases...

I write concrete things, only no one gets it, being weakened by the struggle with the market.

Specific things are research drawings and graphs. R&D, so to speak...

is it instead of a sine wave?

 
Renat Akhtyamov:

is that instead of a sine wave?

Nah, I haven't gone to another dimension yet - I'm just collecting data....

 
Alexander_K:

Nah, I haven't gone to another dimension yet - I'm just collecting data....

Alexander, forgive me for being curious.

Do you even sometimes look at price charts or only collect digital data for machine processing?