Correlation

FRM
#1

I got this a one of the option for a question in schweser material. Couldn’t understand whats wrong with the statement.

Correlation is not an adequate measure of dependence for ellipitical distributions.

I think Correlation is a adequate measure for dependence for ellipitical distributions. Can some one explain why the above statement is wrong ?

#2

My guess is that the question writer was trying to set up a trick for the candidate. I don’t know what the rest of the question looked like but “dependence” sounds as if it is trying to invoke the concept of causality in this context. Correlation measures association between variables, but does not imply that they are dependent on one another (i.e., correlations could be spurious, in which case there is no logical reason that A moves with B – they just randomly did).

To put it in the context of an oft-heard maxim, “Correlation does not imply causality.”

#3

Thanks Destroyer

#4

Thanks trader

#5

You’re welcome.