Contact | Print

# The R package FatTailsR at R/Rmetrics workshops # An introduction to the R package FatTailsR

## FatTailsR: a R package for Kiener distributions and fat tails in finance

### This package (version 1.7-5 issued on 18 June 2015) is dedicated to:

• Kiener distributions of type I, II, III and IV.
• Two different methods and three functions to estimate the distribution parameters.
• Power hyperbolas, powerhyperbolic functions and their inverses,

### Within one function, regkienerLX, a dataset X is fully described with 3 parameters (symmetric model) or 4 parameters (asymmetric model) among :

• symmetric model "K1" : (m, g, k) and (a=k=w, d=e=0)
• asymmetric model "K2" : (m, g, a, w)
• asymmetric model "K3" : (m, g, k, d)
• asymmetric model "K4" : (m, g, k, e)
where :
• m (mu) is the median
• g (gamma) is the scale parameter, different from the standard deviation
• a (alpha) is the left tail parameter, a Pareto and Karamata exponent
• k (kappa) is a global tail parameter, the harmonic mean of a and w
• w (omega) is the right tail parameter, a Pareto and Karamata exponent
• d (delta) is the distorsion parameter, when a is different from w
• e (epsilon) is the eccentricity parameter, when a is different from w. e = d / k

### The excessive use of the inappropriate Laplace-Gauss normal distribution, favoured by many academic textbooks and old regulations, has led to a systematic under-evaluation of the risk and can explain many market crashes in the last 50 years. It is my deep belief that the use of a more suitable model for fat-tailed distributions, like "K1", "K2", "K3" or "K4" models, will result in a better measure of the market risk and the corresponding capital required to cover it.   # A rare moment

## Philosophy in finance

### "In finance", he explained to the audience, "the most important is the helmet". 