Econometrics: why co-integration is needed - page 27

 
faa1947:
Even what you have posted is stifling.

A lot has been done, it's time to put them together into some sort of digestible system for practical use.
 
yosuf:
A lot has been done, it's time to put them together into some digestible system for practical application.


Everything about practice is without comment.

Ready to discuss the application of the listed libraries that were used for this result. Or anything else, but using packages, preferably R.

 
"Theory without practice is dead and practice without theory is blind."

 

What are the tests for cointegration?

  1. The Dickey-Fuller test.
  2. The Engle-Granger test.
  3. The Johansen test.
  4. ?...

Has anyone come across source code in any programming language? Not strong in mathematical formulas. ))

 

https://sites.google.com/site/prof7bit/r-for-metatrader-4/trend-o-mat-arb-o-mat

The guy from forexfactory made a bundle of MT4 with R (at the link), he also wrote two "advisors" that draw a return and a trend synthetic.

There is also a topic on forexfactory where they added also trading on deviations to EAs, I do not remember the exact link :( I think it was written by dirtybrown

 
tol64:

What are the tests for cointegration?

  1. The Dickey-Fuller test.
  2. The Engle-Granger test.
  3. The Johansen test.
  4. ?...

Has anyone come across source code in any programming language? Not strong in mathematical formulas. ))

Everything is in R.
Here moved the wrapper. Downloaded 628

Here moved an example. Downloaded 1047.

You won't be the first.

Clench your teeth and start chipping away at R. Everything will pay off as everything you need is in one place, all docked together, no mismatches and full access from MT4. By picking up the right packages you will get a complete outlook. In particular you will get ready to use the mentioned codes. On R.

 
zkogan:

...

faa1947:

...

Thank you. But I would like to do cointegration tests in MT5. I don't need anything else yet. I don't have any special hopes. It's just a research tool for me.

I have R 2.15.3, EViews 7, MatLab R2012b packages installed. Great programs, especially MatLab. I study even just for warming up and development. But I would like to program all necessary tools in MQL5.

I'm just curious to understand what's going on inside the formulas. )) Explain step by step what operations need to be done to get the cointegration coefficient of two time series. There are many different articles on the internet on this subject, but hopefully they can explain it more clearly here. I'm not interested in measuring, which any mathematical package can do, but in calculating.

A rather detailed explanation can be found in this article: Fundamentals of Statistical Arbitration. Co-integration.

Up to this point:

OK. Suppose we know that two processes are cointegrated. But what does this give us, what mathematical model can be used to represent their dynamics?

...everything is clear, and then I still don't understand how to get the coefficient. For example:

So we arrive at the following error correction model (ECM model):


dY1 = -a1*S + lagged(dY1, dY2)
dY2 = -a2*S + lagged(dY1, dY2)

Where the two variables a1 and a2 come from I still don't understand. And whatlagged(dY1, dY2) means too. ))

 
There are no fish here.
 
yosuf:
A lot has been done, it's time to put them together into some digestible system for practical application.


There are no traders in this thread, only theoretical mathematicians engaged in blatant k-eye))
 
marker:
There are no fish here.
No so no. The question is how to get a co-integration factor, not how to get fish based on the co-integration factor. )))