The interaction of markets. - page 6
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
Очень просто, так как корреляция это следствие, а не причина, то на основании этого можно сделать предположение (вероятность предположения можно увязать со значением корреляции), что если какой-то из двух активов начал движение, то и второй последует за ним.
You make a good point, but the problem is that the second may not follow the first, and the correlation will still be restored. And then, in most cases, we don't know which tool is the "master" and which is the "slave". If you know them, please give us an example. I don't think you can, as there are none, except for the underlying instruments and derivatives for them, but arbitrage on them is impossible for ordinary traders. Checked it out practically.
And in general, if the correlation is a consequence of a linear relationship based on fundamental causes, then it is possible to work on such pairs. But if we get correlation coeff. r -> 1, but there is no fundamental reason, it does not give us the right to do stat arbitrage. Here I agree with you. Here someone wrote that if the euro goes up, so does the pound. What's the fucking correlation here? Two completely different currencies... Or gold to silver... where is it here? Or the S&P_500 vs Dow_Jones... The shares in these indices in relative values of structure (total volume) are the same, but! they are different instruments and they are traded on different exchanges. So, in my opinion, by opening opposite positions in any two (even strongly correlated) instruments, you are also taking a big risk. And it is not arbitrage.
Risk, if you know something about arbitrage, it would be nice to hear more constructive from you. Or links to the source. Stupid boorishness has no respect.
You make a good point, but the problem is that the second one may not follow the first one, and the correlation will still be restored. And then, in most cases we don't know which instrument is the "master" and which is the "slave" . If you know them, please give us an example. I don't think you can, as there are none, except for the underlying instruments and derivatives for them, but arbitrage on them is impossible for ordinary traders. Checked practically.
And in general, if the correlation is a consequence of a linear relationship based on fundamental causes, then it is possible to work on such pairs. But if we get correlation coeff. r -> 1, but there is no fundamental reason, it doesn't give us the right to do stat arbitrage. Here I agree with you. Here someone wrote that if the euro goes up, so does the pound. What's the fucking correlation here? Two completely different currencies... Or gold to silver... where is it here? Or the S&P_500 vs Dow_Jones... The shares in these indices in relative values of structure (total volume) are the same, but! they are different instruments and they are traded on different exchanges. So, in my opinion, by opening opposite positions in any two (even strongly correlated) instruments, you are also taking a big risk. And it is not arbitrage.
Risk, if you know something about arbitrage, it would be nice to hear more constructive from you. Or links to the source. Stupid rudeness has no respect.
Look at the thing, there's a ridiculous premise about "master" and "slave" - Where did I write that?
Next - "If you know of such, please give me an example. I don't think you can" - curtain call.
And what do I have to explain here - that somewhere out there you misunderstood something to me that I can't do it ? (what did I say? I don't understand it myself :)), but in the same spirit)
And then there's this weird blah blah blah ... and what am I supposed to think of you if you have a correlation > 1 ? Tell me what? What should I call you?
"Or the same S&P_500 vs Dow_Jones... the stocks in those indices are the same in relative structure values (total volume)" - relative structure values, structure values, THERE IS NO BREAD ... there is a notion of weight in an index, and the weights of the same stocks in these indices are different .
and then ... "And it's not arbitrage" :))))))))))))))) YES YASIN PEN, where did I claim it was arbitrage ?
What was that all about ? ;)))))
Смотри какая штука, идет нелепый посыл про "ведущий" и "ведомый" - Где я это писал ?
Дальше - "Если ты знаешь такие, пожалуйства, приведи пример. Думаю, ты не сможешь" - занавес.
И что я должен тут объяснять - то что где-то там ты что-то неправильно понял перекладываешь на меня что я не могу это сделать ? (хм, че я сказал ? сам не понял :)), но в том же духе)
Ну и дальше непонятное бла бла бла ... сам с собою, и что я должен о тебе подумать если у тебя корреляция > 1 ? Вот скажи, что ? Как тебя назвать ?
" Или тот же S&P_500 vs Dow_Jones ... акции в этих индексах в относительных величинах структуры (всего объема) одинаковые" - относительных величинах структуры, величинах структуры, ТИХИЙ БРЕД ... есть понятие веса в индексе, и веса одних и тех же акций в этих индексах разный.
и потом ... "И никакой это не арбитраж" :))))))))))))))) ДА ЯСЕН ПЕНЬ, где я утверждал что это арбитраж ?
Что это было вообще ? ;)))))
Look at the thing, there's a ridiculous message going around about "master" and "slave" - Where did I write that?
Your words :
"Do you know what correlation is ?
Do you distinguish between "Can I make an assumption" and "assert" ?
I don't care where a tool goes because I know where the other one goes after it."
Uh-huh... So for you, there's the former and there's the latter. But you just don't know what a master and a slave are... but don't play dumb like you don't understand me and that's my point.
About correlation, between the letter r and 1 I wrote a sign tends (->), read carefully.
In statistics there is such a concept, Relative Structure Value. RIAs describe the fractions, the specific weights of the constituent elements in the total. They are usually obtained in the form of percentages. This is the weight in the index. You're just not familiar with the literature... You're talking to me about indices, I've eaten my teeth on them...
yes I did, where did I say it was arbitrage?
Oh, those woeful arbitrageurs have woken up again :) Have they not all pissed away their spreads yet?
I thought you were a smart man. In other threads somewhere I've read your clever thoughts. But since there is no constructive feedback from you here, you don't have to answer. Time is too valuable. It took me 10 minutes to write you back.
And in general I'm just amazed at the stupidity of most members of this forum (this doesn't apply to Risk).
All Risk have closed the thread.
What does arbitrage have to do with it at all. For example, take gold and silver. It seems to me that there is a fundamental correlation between them because they are two precious metals that in a sense are interchangeable and probably there is a certain average ratio between the price of gold and silver that the market tends to support. Or oil of different grades. That's the kind of correlation I'm talking about. And similar ones.
Look at the thing, there's a ridiculous message going around about "master" and "slave" - Where did I write that?
Your words :
"Do you know what correlation is ?
Do you distinguish between "Can I make a guess" and "assert" ?
I don't care where an instrument goes because I know where the other one goes after it."
Uh-huh... So for you, there's the former and there's the latter. But you just don't know what a master and a slave are... but don't play dumb like you don't understand me and that's my point.
About correlation, between the letter r and 1 I wrote a sign tends (->), read carefully.
In statistics there is such a concept, Relative Structure Value. RIAs describe the shares, the specific weights of the constituent elements in the total. They are usually obtained in the form of percentages. This is the weight in the index. You're just not familiar with the literature... You're talking to me about indices, I've eaten my teeth on them...
yes I did, where did I say it was arbitrage?
Oh, those woeful arbitrageurs have woken up again :) Didn't they all lose out on their spreads yet ?
I thought you were a smart man. In other threads somewhere I've read your clever thoughts. But since there is no constructive feedback from you here, you don't have to answer. Time is too valuable. It took me 10 minutes to write you back.
And in general I'm just amazed at the stupidity of most members of this forum (this doesn't apply to Risk).
All Risk have closed the thread.
What always pisses me off is statements - BUT YOU JUST DON'T KNOW, in the context of YOUR IDIOTIC IMMUNICATIONS. It is only in your head that there are HIGH and HIGH in the stock market, it is only in your head that the correlation tends to something when it is a time-independent value !, it is only in your head that the share of stocks in different indices is the same.
Index expert, shithead. You'd better not get involved and sit quietly as usual, but you got in ... ...and you cleared up all the doubts.
Here is your last chance to justify yourself: What are the two fundamental differences in the calculation of the DOW index from the SP500 ?
In general, it is better to really close the subject, because the gaps and misconceptions are much more serious than I could have imagined.
Меня всегда бесят утверждения - НО ТЫ ПРОСТО НЕ ЗНАЕШЬ, в контексте ТВОИХ ИДИОТСКИХ УМОЗАКЛЮЧЕНИЙ. Это только в твоей башке есть ВЕДОМЫЕ И ВЕДУЩИЕ на фондовом рынке, это только у тебя корреляция к чему-то стремится, когда это величина не зависит от времени !, это только у тебя доля акций в разных индексах ОДИНАКОВАЯ.
Everything is correct, the correlation does not tend to anything, but it is not a constant and depends on time. In the stock market there may be "slaves" and "masters", but the lag dependence will be weak, but it exists: it is detected with the VAR (Vector Autoregressions Model) - Granger Causality.
If they were, I would have been a billionaire a long time ago.
Do you distinguish between: const, f(x1, x2, ...), f(x1, x2, ... t) - ?
Если бы они были, я бы уже давно был бы миллиардером.
Вы отличаете понятия: const, f(x1, x2, ...), f(x1, x2, ... t) - ?
One more time - correlation depends on time, that's why it changes. If had but for what - it is my personal experience, lag dependence results up to 30%, it is a fact, it is written about it even in smart books. Don't worry, you cannot know everything - you just have to admit the error.
What the hell kind of mistake?
It's about correlation and full stop, if you don't know what it depends on and what tends to be, that's your little problem.
If there is a lag correlation and you are not a millionaire, then you are a WRUNG !
"My personal experience ... lag dependence up to 30%" - idiot, lag is actually expressed in units of time. What the hell are you counting? ;))))
Какую нахрен ошибку ?
Шла речь о корреляции и точка, если ты не знаешь от чего она зависит и что такое стремится - это твои маленькие проблемы.
Если есть лаговая зависимость и ты не миллионер, значит ты ВРУН !
Yeah, and you wrote that the correlation does not tend to anything, but here's a miracle - it does not depend on time. If it were so, the correlation would tend to a certain value.
And don't go off-topic: we were also talking about "masters" - "slaves". Why do I need 30% of I(1) when I have 99.9% of I(0).