How to WEIGHT the adjusted Fisher–Pearson standardized moment coefficient (skewness).

#1
Hello,

I'm using the adjusted Fisher–Pearson standardized moment coefficient formula to calculate the skew of some sample data (it's the same formula that uses Excel - skew()).

The problem it's that I need to weight the points in the sample data; because the data should be time-weighted. This is: the more recently the data has been collected, the more influence it should have in determining the skewness of the sample.

The weighting scheme should be as simple as a constant increment by time elapsed period: 1*Constant*X(1), 2*Constant*X(2), ..., N*Constant*X(N), where X(1) is older data than X(2), and so on.

I will GREATLY appreciate the help of anyone who could chime in and give me some clue about how to solve this. Or point me to the right direction (some paper, etc.) because this is not my field and I don't know the literature.

Thanks in advance.
 

Englund

TS Contributor
#2
Re: How to WEIGHT the adjusted Fisher–Pearson standardized moment coefficient (skewne

I guess a GARCH model is feasible in this case. Not sure an estimate of the skewness can be derived using a GARCH model though, but it should since it utilizes a normally distributed variable.
 
#3
Re: How to WEIGHT the adjusted Fisher–Pearson standardized moment coefficient (skewne

Thanks for the pointer Englund, I'll take a look at GARCH model.

Sorry for not saying thanks before, I don't have too much time to check the board frequently.