Corrado Provasi

Simulation and Estimation of the Meixner Distribution

The Meixner distribution is a special case of the generalized z-distributions. Its properties make it potentially very useful in modeling short-term financial returns. This article proposes an algorithm to simulate the Meixner distribution, and shows how to obtain maximum likelihood estimators of its parameters. A GARCH-type model is then assessed, assuming that the innovation distribution is a standardized Meixner. Goodness-of-fit properties are investigated for some real financial time series, using bootstrap tests based on the empirical process of the residuals.

Confidence Interval Estimation for Inequality Indices of the GiniFamily

In this paper we present some nonparametric bootstrap methods to constructdistribution-free confidence intervals for inequality indices belonging to theGini family. These methods have a coverage accuracy better than that obtainedwith the asymptotic distribution of their natural estimators, typically thestandard normal. The coverage performances of these methods are evaluated forthe index R by Gini with a Monte Carlo experiment on samples simulated fromthe Dagum income model (Type I), which is usually used to describe the incomedistribution.

Approximated Distributions of Sampling Inequality Indices

Often, in finite samples, the true level of the confidence intervals for natural estimators of inequality indices belonging to the Gini family differs greatly from their nominal level, which is based on the asymptotic confidence limits. This paper shows how the Gram-Charlier series can be used to obtain improved finite-sample confidence intervals. Our work focuses on the implementation in Mathematica 3.0 of computational procedures to compute the Gram-Charlier distribution for the following sampling inequality indices: R by Gini, P by Piesch and M by Mehran for the Dagum (Type I) distribution. The results of a Monte Carlo experiment confirm that, for the cases investigated, the Gram-Charlier distribution largely eliminates the problem of incorrect finite-sample level.

Misspecification Testing for the Conditional Distribution Model in GARCH-Type Processes

In this article, we study goodness of fit tests for some distributions of the innovations which are usually adopted to explain the behavior of financial time series. Inference is developed in the context of GARCH-type models. Functional bootstrap tests are employed, assuming that the conditional means and variances of the model are correctly specified. The performances of the functional tests are assessed with a Monte Carlo experiment, based on some of the most common distributions adopted in the financial framework. The results of an application to the series of squared residuals from a PARCH(1,1) model fitted to a series of foreign exchange rates returns are also shown.

Nonparametric Estimation Methods for Sparse Contingency Tables

The problems related with multinomial sparse data analysis have been widely underlined in statistical literature in recent years. Concerning the estimation of the mass distribution, it has been widely spread the usage of nonparametric methods, particularly in the framework of ordinal variables. The aim of this paper is to evaluate the performance of kernel estimators in the framework of sparse contingency tables with ordinal variables comparing them with alternative methodologies. Moreover, an approach to estimate the mass distribution nominal variables based on a kernel estimator is proposed. At the end a case study in actuarial field is presented.

GSH Dependence Modeling with an Application to Risk Management

The generalized secant hyperbolic distribution (GSH) can be used to represent financial data with heavy tails as an alternative to the Student-t, because it guarantees the existence of all moments, also with a high kurtosis value. In order to obtain a multivariate extension of the GSH distribution, in this article we present two approaches to model the dependence, the copula approach and independent component analysis. Since the methodologies considered allow to simulate the GSH dependence, we show also the empirical results obtained in the estimation of risk of a financial portfolio by the Monte Carlo method.

Design and selection of products via genetic algorithms and neural networks

In the paper we address the design and selection of products in the framework of genetic algorithms and neural networks. We propose a procedure in alternative to the current methodologies and illustrate its advantages and disadvantages. The results of a real‐world application are reported. Although these results refer to a specific consumer product, we believe that it is worth to continue the analysis of this methodology. Copyright