Coskewness & Cokurtosis- How relevant are these to the exam?

[deji]

I'm new to BT and the FRM.

I'm reading through the GARP material and it keeps mentioningthat coskewness and cokurtosis aren't widely applied in industry. I'm a little behind where I want to be with my preparation and the explanations are quite complicated and hard to follow. Can I skip this section?

Thanks,

Deji

 

[David Harper CFA FRM]

Hi Deji, I don't think there is a knowable answer, with any certainty, to your question. (the terms coskewness and cokurtosis have never appeared before, they are entirely new in "Interpret the skewness and kurtosis of a statistical distribution, and interpret the concepts of coskewness and cokurtosis). The FRM has traditionally over-assigned the math on skew and kurtosis so history suggest low testability but, on the other hand, these are one of a few uniquely new terms in stats. Thanks,

 

[deji]

Thanks David. I'll stick a pin in it and hopefully be able to come back to it later.

 

[JulioFRM]

have these concepts been tested after 2013?

 

[David Harper CFA FRM]

Hi @JulioFRM I am not personally aware of coskew or cokurtosis having ever been tested, is just my perception. While they are in an LO "Interpret the skewness and kurtosis of a statistical distribution, and interpret the concepts of coskewness
and cokurtosis," they are very low testability. I've not seen a GARP practice question mention them, either. FWIW.

 

[Flashback]

Are the coskew and cokurtosis nothing else than covariances of skewed or distribution with deviated kurtosis?
I mean, we should just interpret them.

 

[JulioFRM]

Another person explained these concepts very nicely as follows:

"Coskweness measures the skewness of one variable with respect to the other variable, while in the skewness we measure whether the variable is positively or negatively skewed, in coskewness we find the relative skewness that is whether the variable is positively or negatively skewed with respect to the other variable. Similarly the Coskurtosis measures the peakedness of one variable with respect to the other variable, while in the kurtosis we measure whether the variable is peaked(over>3 or under<3 kurtosis), in Coskurtosis we find the relative kurtosis that is whether the variable is over or under peaked with respect to the other variable."

source: https://forum.bionicturtle.com/threads/coskewness-and-cokurtosis.9365/#post-40428

 

[Flashback]

OK. Thank You, Julio.