From theory to practice - page 99

 
Maxim Dmitrievsky:

Non-linear price is correspondingly nonlinear also time, this was shown by Mandelbrot in the examples with the multiplicative cascade


Greetings Maxim! The tick arrival time is non-linear - yes. Moreover, it is not exponential, although it tends to be. But, the price... or is it a price scale after all? I'd better go and read.

 
Maxim Dmitrievsky:

The price is non-linear, so is time, as Mandelbrot showed with the examples of the multiplicative cascade.


Price changes relative to time otherwise it would be constant.

 
Uladzimir Izerski:

Price changes relative to time otherwise it would be constant.

And the price scale? It never even occurred to me that it could be non-linear. Or did I misunderstand the question?
 
Alexander_K2:

Greetings Maxim! The tick arrival time is non-linear - yes. Moreover, it is not exponential, although it tends to be. But, here's the price... or is it a price scale after all? I'm off to read.


Goodnight :) I don't know, it's a bit different terminology. It's better to read the primary source, so there's no mistakes like with a deaf phone.

and whether this great knowledge will help in trading is also a question.

 
Uladzimir Izerski:

Price changes relative to time or it would be constant.


and time in the market is not constant, that's what I mean.

So it goes up and down

 
Maxim Dmitrievsky:

Goodnight :) I don't know, it's a bit of a different terminology there. It's better to read the primary source yes, so that there are no mistakes like with a deaf phone

and whether this great knowledge will help in trading is also a question

I, for example, have benefited greatly from understanding that tick arrival times are non-exponential, but very close to it.

That's why I myself am forced to work at exponential frequency - the smartest ticks filter has turned out.

I think any knowledge is important - you just have to figure out how to use it in practice

 
Alexander_K2:
And the price scale? It hadn't even occurred to me that it may be non-linear. Or maybe I misunderstood the question?

If time doesn't change, the price doesn't change.

Maxim Dmitrievsky:

And time in the market is not constant, that's what I mean.

So it goes up and down.


As time changes, so does the speed of price change.

 

By and large, as the discussion progressed, I realised that reading tick data at exponential intervals was the only correct way to collect data. The histograms I'm posting here are proof of that. Any other method results in some kind of muddle, and pure Steward shows up that way.

 
Alexander_K2:

I, for example, find it very helpful to understand that the tick arrival time is non-exponential, but very close to it.

That's why I myself am forced to work at an exponential frequency - the most gorgeous tick filter has turned out.

I think any knowledge is important - you just have to figure out how to use it in practice


I'm curious. You analyze ticks, while the result is calculated as "days" or "hours". Is the time spent on such an analysis justified?

 
Alexander_K2:

I, for example, find it very helpful to understand that the tick arrival time is non-exponential, but very close to it.

That's why I myself am forced to work at an exponential frequency - the most gorgeous tick filter has turned out.

I think any knowledge is important - you just need to figure out how to use it in practice


Yes, I understand a little of what you mean, but I haven't really thought about it yet... so as usual, good luck :)

Reason: