Modelling the extreme variability of the US Consumer Price Index inflation with a stable non-symmetric distribution

Abstract

Stable distributions have interesting properties that make them a versatile tool suitable for modelling a wide range of processes from different scientific fields, from meteorology to computer science and from communications to economic theory. Our objective is to use stable laws to get an insight at the distributional characteristics and behavior of the US Consumer Price Index inflation. Such a descriptive model is essentially an easy to use tool that provides us with useful information about the Index, via its ability to generate series with similar characteristics. Besides using an appropriate non-parametric test, an examination via an ensemble of a large number of simulated series is implemented in order to assess the accuracy of the model. Its capabilities and adaptability make it a useful tool for everyone analyzing processes from the field of economics.