Econometrics: let's discuss the CU balance sheet. - page 3

 
faa1947:

I don't get it. SD calculated before detrending or after?


Forget about "detrending" - no one is interested in it when analysing trade results

the system yield for the period is divided by the average standard deviation of the TC yield for the same period

The daily standard deviation is determined by dividing the daily percentage change in the play money account by the state of the account at the beginning of the period

 
49 deals I didn't even notice, not enough, of course
 
Demi:

Forget about "detrending" - no one is interested in it when analysing trade results

No, detrending is just to cut off the random results obtained by using this trend. I.e. what I wrote in a previous post, or what Taleb wrote - "do not confuse the bull market with your own genius". But you can of course do without it. And it doesn't save from all accidents.
 
Avals:

no detrending is needed just to cut off the random results obtained by using this trend. I.e. what I wrote about in the previous post, or what Taleb wrote about - "don't confuse a bull market with your own insignificance". But you can of course do without it. And it doesn't save against all accidents.


why do you need "detrending" to analyse the results of the TS over a certain period? There is a balance graph - WHY do we need "detrending" ?????????????? What for?

How can you use a "trend" on a balance sheet chart?

 
Demi:

Why do you need "detrending" to analyze the results of TC for a certain period? You have a balance graph - WHY do you need "detrending" ?????????????? What for?


I wrote in my previous post why econometricians do it. It can be done otherwise, without detrending, but apparently it is easier for them)).

P.S. not on the balance chart of course, but on the row itself

And so, probably that's how faa analyzed the error

 
Avals:

I wrote in a previous post why econometricians do it. It could be done differently, without detrending, but it's probably easier for them.)


Which econometricians use "detrending" the balance sheet graph to analyse TC efficiency? Taleb doesn't have it - don't talk shit about him. Who wrote it? Where?

 
Demi:


Which econometricians apply the "detrending" of the balance sheet graph to analyse TC efficiency? Taleb doesn't have it - don't talk tough on him. Who wrote it? Where?


Detrending the series, not the balance sheet graph. He detrended the balance in order to analyse the error
 
Something is not working with a large number of trades. I'm taking time out and will post the result as soon as I find the reason
 
Demi:


Why do we need "detrending" to analyze the results of TS for a certain period? There is a balance graph - why do we need "detrending" ?????????????? What for?

How can you use a "trend" on a balance sheet chart???


Consider specifically. Here is a year's worth of EURUSD.

Here are the statistics for this sample.

SD = 522 pips.

Detrending.

Here are the statistics for the noise. This is pure noise, unaffected by the deterministic component.

The SD is already 18 pips. The difference with 522 pips was given by deterministic component.

We take the part of the sample which corresponds to the trend at the beginning.

Statistics for this sample without detrending.

SD = 349 pips. It is clear, as a smoother part was selected.

Let's detrend.

Statistics for noise:

SD = 14 pips. Not very different from the whole sample.

Any quotes are non-stationary series due to trends. so SD is not applicable to them. which is shown in the figures.

If we analyse noise = kotir - smoothing, there is no trend there and if the series remains non-stationary, it does not have a deterministic component clogging the statistics.

Conclusion: SD for quotes without detrending is just a numbers game.

 
faa1947:


Consider concretely. Here is a year's worth of EURUSD.

Here are the statistics for this sample.

SD = 522 pips.

Detrending.

Here are the statistics for the noise. This is pure noise that is not affected by the deterministic component.

The SD is already 18 pips. The difference with 522 pips was given by deterministic component.

We take the part of the sample which corresponds to the trend at the beginning.

Statistics for this sample without detrending.

SD = 349 pips. It is clear, as a smoother part was selected.

Let's detrend.

Statistics for noise:

SD = 14 pips. Not very different from the whole sample.

Any quotes are non-stationary series due to trends. so SD is not applicable to them. which is what the figures show.

If we analyse noise = kotir - smoothing, there is no trend there and if the series remains non-stationary, it does not have a deterministic component clogging the statistics.

Conclusion: SD for quotes without determinism is just a numbers game.


That's all I understood.

Did you understand my question - why "detrending" is necessary? to analyze results of TS for a specified period? There's balance chart - WHY make a "detrending"????????????? What for?