FAT Tail Returns on Hedge Funds

[broadex]

I understand returns from hedge funds are negatively skewed and have excess positive kurtosis. What i cant figure out (even after reading a number of threads here) is why excess positive kurtosis is considered fat tails. This sounds counter intuitive given a tall/peaky distribution will have thinner tails.

[whatsyourgovt]

I think this is the least of your worries but just remember it is leptokurtic and thus sharper peak, fat tails.

[broadex]

Thanks Whatsyourgovt. If this comes up in the exam i will get the answer right but i just wanted to understand the rational…

In addition i would appreciate if someone explain why positive kurtosis is considered riskier. For instance we say hedge funds have positive kurtosis ie most of distributions around the meanyet this is considered risky ???

[Iconoclasts]

@broadex

Thanks Whatsyourgovt. If this comes up in the exam i will get the answer right but i just wanted to understand the rational…

In addition i would appreciate if someone explain why positive kurtosis is considered riskier. For instance we say hedge funds have positive kurtosis ie most of distributions around the meanyet this is considered risky ???

Sure it’s more peaked, but you also have more extreme events than predicted by a normal distrubution. A larger % of large losses (and wins) would be considered more risky, right?

[broadex]

@Iconoclasts

@broadex

Thanks Whatsyourgovt. If this comes up in the exam i will get the answer right but i just wanted to understand the rational…

In addition i would appreciate if someone explain why positive kurtosis is considered riskier. For instance we say hedge funds have positive kurtosis ie most of distributions around the meanyet this is considered risky ???

Thanks but if the the distribution is more around the mean ( morepeakedness -if there is such a term) than a normal distrubution then intuitively the chances of having an outlier should be less for a positive kurtosis than a normal distribution = less risky (intuitively). I know the answer is the opposite but i cant figure out a way of explaining this.

[S2000magician]

Leptokurtic distributions tend to have higher probabilities of returns near the mean (say μ ± 0.5σ), lower probabilities in the midrange from the mean (say, between μ + σ and μ + 2σ, and between μ – 2σ and μ – σ, and higher probabilities at the extremes (say, above μ + 3σ, and below μ – 3σ).

[broadex]

@S2000magician

Leptokurtic distributions tend to have higher probabilities of returns near the mean (say μ ± 0.5σ), lower probabilities in the midrange from the mean (say, between μ + σ and μ + 2σ, and between μ – 2σ and μ – σ, and higher probabilities at the extremes (say, above μ + 3σ, and below μ – 3σ).

Good explanation …thanks Magician!