Not the Grail, just a regular one - Bablokos!!! - page 55
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Hm.
What is a regression?
For example, EURUSD = a+ b*GBPUSD
are evaluated by a and b - that's the coefficients for you. So regression has a lot to do with it.
Look here. There is a lot on this subject.
Well in principle we are interested in only one factor b. For the resulting series EURUSD-b*GBPUSD to be stationary. Only you have so many reports and the coefficient is different everywhere, even if the sample size is not much different. I.e., after opening positions you will have to constantly correct their volumes. This is an additional cost that can eat up all the potential profit.
You don't seem to understand the meaning of the word cointegration.
My participation in the forum has a completely mercantile interest: pair trading, apart from coding, is a constant and tedious job. It seemed to me that you have no relevant training. Or am I wrong?
You don't seem to understand the meaning of the word cointegration.
My participation in the forum has a completely mercantile interest: pair trading, apart from coding, is a constant and tedious job. It seemed to me that you have no relevant training. Or am I wrong?
I am aware of what cointegration is. It's just that you stubbornly don't want to understand that your found cointegration is simply an adjustment to a certain time interval. And beyond that interval there is most likely no more cointegration (at the same coefficients). I'm not an expert in econometrics etc, but I've read that there is such a thing as a false regression, and it can result in making the wrong assumption about a relationship where there really is none.
Initial quotes, H1 6736 bars, year.
Residuals graph from cointegration regression
Testing for non-stationarity. The numbers on the left are, roughly, the probability that the residual is non-stationary.
One of the outliers = 2.4% is the probability that the residual is non-stationary.
Exactly, we cannot reject the null hypothesis that the series is stationary at nearly 100%!
Nice drawings. Just epochally stationary residual turned out))
Have you tried to output the residue of a regression for example EUR/USD on the same EUR/USD but shifted by some reporting period... month... or quarter?
I am aware of what cointegration is. It's just that you stubbornly do not want to understand that your found cointegration is just a fitting for a certain time interval. And beyond that interval there is most likely no more cointegration (at the same coefficients). I'm certainly no expert in econometrics, etc., but I've read that there is such a thing as a false regression, and as a result one can make the wrong assumption of a relationship where there really isn't one.
Let's, once again, how the figures are derived. a sample of 6736 bars is taken. a window = 236 bars is taken. Move this window from left to right. calculate the unit root test and write in the outermost, 236, place. We get 6736-236 measurements. We see a variable value of the unit root test. The coefficients naturally change.
So what is the question?
Nice drawings. It's just an epochal stationary residue)))
Have you tried to show the residue of e.g. EUR/USD regression to the same EUR/USD but shifted by some reporting period... month... or quarter?
What is "regression EUR/USD on the sameEUR/USD"? Could we have a formula?
I don't understand, why a shift?
What is a "regression of EUR/USD on the same EUR/USD"? Couldn't we have a formula?
I don't understand, why shift?
I just think it is possible to have a good correlation when overlaying the current graph of the previous reporting period. ... although I'm not very good at it... sorry for getting stuck in))
It's called an autocorrelation function. It is very widely used.
AA. I see.
There's a reason, they must have invented a shift in the dummy indicator by a certain number of bars... - so the autocorrelation can be calculated.)
AA. I see.
There's a reason, they must have invented a shift in the dummy indicator by a certain number of bars... - so the autocorrelation can be counted.)
Oh, come on, Lizavetta.
You're a professional.
The ACF begins with Box and Jenkins. What's that got to do with the mash-ups? It's for RBC's solidity.