Arbitration Advisor - page 9
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
Isn't there anyone using co-integrated currency pairs?
CoNegration looks great on the story))
What window interval do you use, on what timeframe, how many pairs, what method of selecting pairs, what method of getting the spread?
i would be interested to see the implementation of the paired filter (not smoothing) calman
indicator for that, but without stationarity tests
https://www.mql5.com/ru/code/11859
Co-integration looks good on history))
what interval do you take the windows, on what timeframe, how many pairs, what is the method of getting the spread?
I'd be interested to see the incarnation of the paired filter kalman
the indicator is for that, but without stationarity tests
https://www.mql5.com/ru/code/11859
Co-integration is making decisions on a stationary series with non-stationary quotes as input.
Then in the future the results will be the same as on history
Co-integration is making decisions on a stationary series with non-stationary quotients as input.
Then in the future the results will be the same as in the history
will not)
and you haven't shared the answers to my questions)
OK, let's have a few pictures as an introduction
won't)
And you didn't share the answers to my questions)
I have no answers to your questions.
And I am interested in the experience of the classics of this question. I have such an EA with a profitability of up to 5 pips per position.
PS.
Kalman is not needed. It is used in other models
I have answers to your questions.
And I am interested in the experience of the classics of this question. I have such an EA with a profit margin of up to 5 pips per position.
PS.
Kalman is not needed. It is used in other models
Ok, let's have a few pictures for refresher
I think the orthogonal regression is not applicable.
The regressions should be such that the residual is stationary. You get a residual that is NOT stationary - the ADF test fails - no further work with such a residual makes sense.
PS.
Which package are the pictures from?
I think orthogonal regression is not applicable.
The regressions should be such that the residual is stationary.
PS.
Which package are the pictures from?
San sanych, you are way behind the times.
the Kalman filter has been used for a long time. you can read about its use in Ernie Chan. there's a code for Matlab there. and such a model is in the link below
The orthogonal fits better for one reason - it is symmetric, unlike the usual OLS (if you swap pairs, your regression coefficient will change)
https://www.pairtradinglab.com/analyses/WDsreX1v6miBQ59k
san sanich, you are out of touch.
the Kalman filter is applied and has been for a long time. you can read about its application in Ernie Chan. there is a code for matlab there. and such a model is in the link below
The orthogonal fits better for one reason - it is symmetric, unlike the usual ISC (if you swap pairs, your regression coefficient will change)
https://www.pairtradinglab.com/analyses/WDsreX1v6miBQ59k
On Kalman, I'm not lagging behind. I'm well aware of where it applies, and I also know that in co-integration Kalman does not apply.
I gave my opinion on orthogonal based on your results. You need a stationary residual. Other regressions are used for that.
On Kalman, I am not behind. I am well aware of where it applies, and I also know that Kalman does not apply in co-integration.
I gave my opinion on orthogonal based on your results. You need a stationary residual. Other regressions are used for that.
You say other models, other regressions, a stationary residual.
Pruf and examples - which models, which regressions. what's the advantage?
Otherwise the dialogue turns out to be without constructiveness and without your examples.
If the question of whether someone uses cointegration was idle, well, then we can stop here.