Leptokurtosis


Also found in: Medical, Encyclopedia.

Leptokurtosis

The condition of a probability density curve to have fatter tails and a higher peak at the mean than the normal distribution.

Leptokurtosis

A state in which the volatility of a security is itself not volatile. That is, lepkurtosis is a state in which the volatility of a security changes at a relatively low rate. This is shown on a chart by a distribution line with data points resembling fat tails and a higher mean, with an even distribution. See also: Kurtosis, Platykurtosis.
References in periodicals archive ?
Hence, the contracts show existence of leptokurtosis in the mean data generating process, with too many observations around the mean and in the both tails and too few observations around one standard deviation from the mean, seem present.
Second, leptokurtosis in the unconditional distribution of output growth vanishes after adjustment for (G)ARCH with conditional normality.
to the habitat as a factor enhancing leptokurtosis and culminating with
Since then, numerous studies have demonstrated that market returns are not normal, but instead exhibit leptokurtosis (try this term at a cocktail party
The conditional heteroscedasticity captures leptokurtosis in the unconditional distribution.
For the purpose of evaluating the degree of leptokurtosis, that is, the prevalence of fat tails in the conditional volatility patterns, we apply the generalized error distribution (GED) parameterization.
VaR is a best guess subject to attendant assumptions and complications as represented by skewness, leptokurtosis and outliers (especially), and conditional volatility and conditional correlation.
The GARCH-M process reduces the sample kurtosis, but fails to fully account for leptokurtosis.
The results suggest that all data series are skew, exhibit excess leptokurtosis, or both, that is, they have fat tails and/or a sharper central peak than the standard normal distribution.
Using data from the stock indexes of London, New York, Tokyo, and Toronto, they find that ENN models outperformed many traditional models, including the autoregressive conditioned heteroskedasticity (ARCH) model, in removing leptokurtosis and symmetric and asymmetric heteroskedasticity from the stock index data.