Zero sample correlation does not necessarily mean there is no linear relationship - page 59
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
OK, if one considers the series as a whole, yes. But we can talk about a piecewise constant vector of cointegration. For those two stocks it would be exactly that.
No, because the convergence rate of the hedge ratio estimate is proportional to N, not sqrt(N), where N is the number of observations.
You're a bad telepath :D
I'm bald ))sorry))
Piecewise all forex rows are cointegrated.
well excuse me))
Piecemeal all the rows in the forex are cointegrated.
The fact of cointegration does not explain the following circumstance.
I take some window and count the cointegration vector on it. Then I shift this window by 1 and count the vector again. I graph the first coefficient of the vector.
The cointegration vector is a variable. Thus, co-integration as a phenomenon is always present in my sample, while the vector varies.
Now for the forecast. We enter the market at zero crossing of the residue from the cointegration equation. For this moment there was a vector of cointegration. Then this vector changes, but we actually exit for another vector of cointegration? Is it stationarity? Personally, it confuses me.
The fact of cointegration does not explain the following circumstance.
I take some window and count the cointegration vector on it. Then I shift this window by 1 and count the vector again. I graph the first coefficient of the vector.
The cointegration vector is a variable. Thus, co-integration as a phenomenon is always present in my sample, while the vector varies.
Now for the forecast. We enter the market at intersection of zero residue from the cointegration equation. For this moment there was a vector of cointegration. Then this vector changes, but we actually exit for another vector of cointegration? Is it stationarity? Personally, it confuses me.
I don't understand the point of this option - how do you consider the vector of cointegration? Between which rows? Are you calculating autocorrelation?
Dimitri, you have already written that you "no longer participate in the correlation threads here".
Well, what do you want?
Dimitri, you've already written that you "no longer participate in correlation threads here".
Well, what do you want?
Well, it worked a little bit by inertia... I don't want anything.
Proverb
There was a time when a mighty and wise king ruled a distant city called Virani. His power filled with fear and awe, and his wisdom inspired love. In the centre of Virani stood a well with cool, clear water. All the people of the town and even the king and the court lords drank from it,
as there was no other well.
One night, when everyone was asleep, a witch came to the city and poured seven
drops of a wicked potion into the well, and she said: "From now on, whoever drinks this water will lose his mind!"
In the morning, all the citizens, except the king and his chief adviser, drank of this water, and madness took possession of their minds - the witch's spell had come true. All day long, people whispered to each other in alleyways and market squares: "The tsar has gone mad... Our Tsar and his Councillor have gone mad... "How can we be ruled by a mad tsar! "Off with him from the throne!"
That night, the king ordered a golden cup to be filled with water from the well. And when the cup was brought to him, he took a big gulp and gave it to his advisor to drink. And a riotous merriment spread over the distant city of Virani, for the king and his counsellor had their wits returned to them.
Khalil Jabran. From the book "The Madman. His Parables and Poems."
I apologise for barging into the conversation.
Please enlighten me:
For pair trading what is used ?
Correlation or cointegration
Stells:
What is used for pair trading ?
I apologise for barging into the conversation.
Please enlighten me:
For pair trading what is used ?
Correlation or cointegration ?
Ido not know how to do it:
Co-integration.
That means pairs should be cointegrated at a certain (last) time interval ?
Can you advise me to read more about it (cointegration, steam trading)?
So the pairs have to be cointegrated at a certain (last) interval of time ?
Not necessarily))
Can you advise me to read something on the subject (cointegration, pair trading) ?
Google - definition of cointegration, Engle-Granger and Johansen tests, how to hedge (you can use cointegration vector/make a beta-neutral portfolio, etc.).
http://www.amazon.com/Pairs-Trading-Quantitative-Methods-Analysis/dp/0471460672 - the basics are well laid out.