Econometrics: why co-integration is needed - page 23

New comment
 
Vizard:

So it's also an indicator...just a multi-currency one...but that's not the point...the point is that the TS on it will get better as it optimises... i.e. the same classic approach...
In TA you look for patterns with unknown stat characteristics. It's very close to guessing by coffee grounds. I look for statistical characteristics of rows and on this basis I predict future behavior. For example, to our rams. Within the framework of this approach, the sacred cow "you cannot over-sit" is dead. You can wait too long because you will still get to zero and the loss on the way to zero does not have the property of growing indefinitely.
 
HideYourRichess:

I see, if you take out @TREND we have a simple equation "difference = eurusd - gbpusd * K". Then yes, you play swings, like on a flat, bounce from channel boundaries to zero. Only not on the prices themselves, but on their "difference".

Once again, just in case. Declaimer, excuses. No ratios, no co-integration guarantees that the music will last forever. Due to the very nature of exchange rates. There is no stationarity there in a synthetic instrument (on your trading horizons), no matter what the unit roots tell you about it.

The trend should not be thrown out. Looking at 6700 bars

Clearer sko

And the unit roots

Which shows that we need to work on cointegration and the unit root test works. At any rate, if the residual is not stationary, then you should stay out of the market. Agree that's a lot.

 
faa1947:
In TA you are looking for patterns with unknown statistical characteristics. It is very close to reading coffee grounds. I look for statistical characteristics of rows and on this basis I predict future behaviour. For example, to our rams. Within the framework of this approach, the sacred cow "you cannot over-sit" is dead. You can "oversleep" as you will come to zero anyway and the loss on the way to zero doesn't increase infinitely.


If you're looking for a pattern, everything is very well formalized... if the pattern fits the market - trade... no - on the fence... no guessing...

 
faa1947: In TA you are looking for patterns with unknown statistical characteristics.
This is only a very limited understanding of TA, as its fans or haters insist. A more general understanding includes all econometrics.
 
Mathemat:
This is only an extremely limited understanding of TA, as its fans or haters insist. A more general understanding includes all econometrics.
Which includes what. TA specialists used to be called chartists. It is an extremely advanced means of drawing graphs. No question about it. Everything else, like analysis, forecasting - just questions and nothing else.
 
Vizard:


If the pattern fits this market well, trade...

This is a question that cannot be answered within TA. That is the problem.
 
faa1947:
I look for statistical characteristics of rows and on this basis predict future behaviour. For example, to our rams. Under this approach the sacred cow "you can't overshoot" is dead. It is possible to hibernate, because we will come to zero anyway and the loss on the way to zero does not have the property of growing indefinitely.
The most egregious example that this is not true is the collapse of the LTCM fund. They too, played the spreads, on very stationary synthetics.
 
faa1947:

The trend cannot be discarded. Looking at 6700 bars

Clearer sko

And the unit roots.

Which shows that we need to work on cointegration and the unit root test works. At any rate, if the residual is not stationary, then you should stay out of the market. Agree that this is a lot

Well, what can I say, if you think that cointegration, unit roots, etc. - is cool, that's up to you. My job is to warn you that everything in this subject is not as cool as the books say. You need to understand where the black swans may sit in this methodology.
 
HideYourRichess:
Well, what can I say, if you think that cointegration, unit roots, etc. - is cool, that's up to you. My job is to warn you that everything in this subject is not as cool as the books say. You need to understand where the black swans may sit in this methodology.
Where?
 
HideYourRichess:
The most egregious example that this is not the case is the collapse of the LTCM fund. They too, played the spreads, on very stationary synthetics.
I wouldn't generalise like that. By and large they relied on an efficient market. And when the market had a memory, it all collapsed.
1...16171819202122232425262728
New comment