From theory to practice - page 68

 
Nikolay Demko:

Of course it will be different, that's what the Bolinger is built on.

Bollinger Bunds is the SCO deferred from the MA, subtract the wrist from the BB lines and get the SCO.

There's even in the BB settings how much RMS to defer from the MA.


Incorrect: the discussion here is about increments, not the cotier itself.

It is argued that the tick data increments will have the same mo and variance, which is not the case. Although Asaulenko will surely find areas where this will hold.

 
Максим Дмитриев:

wanted to calculate the standard deviation and found out that this is the same as the cattle.)


Oh, how many wondrous discoveries we have... )

 

The standard deviation, aka sko, aka sigma.



I now know the three sigma rule.

which says thatnormally distributed data have 997 values out of 1000 within ±3*SCO of the arithmetic mean.


 
СанСаныч Фоменко:

Incorrect: the discussion here is about increments, not the quotient itself.

It is argued that the tick data increments will have the same mo and variance, which is not the case. Although Asaulenko will surely find areas where this will hold.


Please check your assertion. Because it does not fit with my checks.

 
Nikolay Demko:

Oh, how many wondrous discoveries we have... )


Yeah)

 

Here is the incremental pava



And here is a graph of the average in window 100, shifted by 1


 
СанСаныч Фоменко:

Here is the incremental pava



And here is a graph of the average in window 100, shifted by 1



Now take the diff from the mach and get the same graph of the average.

If you subtract the dipmeter from the BB band and take the power of diffraction, you will get the RMS curve.

You don't need to take the differential from the RMS chart, because it is derived from the scale. Just subtract the mach mach from the BB and you get the same series as the RMS from the average differential.

 

I wrote at the beginning of this thread that it is generally accepted that financial series are not stationary. In the early 70's this was established in ARMA models and over the last 40 years all development has been to take into account the nuances of this very non-stationarity.

And here is a so-called "physicist" who simply rejects this and pushes the question for 70 pages, taking stationarity as the basis, calculating some statistics, assuming that what he calculates on history will not change in the future.

Given that the issue of non-stationarity of financial series is universally acknowledged, I allow myself to express myself to this very "physicist" rather crudely.

 
СанСаныч Фоменко:

Completely agree with you when you study history.

But the problem is that NOT stationarity will be "outside" where you will be making trading decisions. This is the point of NOT stationarity in that you will be making decisions in areas that have NOTHING to do with the beautiful, stationary history you have studied. And the worst part is that the variance is bound to be multiples (or maybe orders of magnitude) greater than what you get on the stationary plot

SanaSanych, I'm not talking about history, I'm talking about the principles of real-time signal processing. In this case the history is auxiliary and we are more interested in the here and now.
 
Nikolay Demko:

Now take the differential from the mach and get the same graph of the mean.

And if you subtract the waving from the BB band and take the differential, you get the RMS graph.


Agreed.

Your last post is correct, and the previous one not so much, as it assumed the knowledge stated in the last post.

But all this is minor compared to the levels of fakeliness of this thread.

Reason: