In this paper we confront sensitivity analysis with diagnostic testing.Every model is misspecified, but a model is useful if the parameters of interest (the focus) are not sensitive to small perturbations in the underlying assumptions. The study of the e ect of these violations on the focus is...

We propose a class of distribution-free rank-based tests for the null hypothesis of a unit root. This class is indexed by the choice of a reference density g, which needs not coincide with the unknown actual innovation density f. The validity of these tests, in terms of exact finite sample size,...

Let (X1, Y1),…., (Xn, Yn) be an i.i.d. sample from a bivariate distribution function that lies in the max-domain of attraction of<br/>an extreme value distribution. The asymptotic joint distribution of the standardized component-wise maxima<br/>√n i=1 Xi and √n i=1 Yi is then characterized by the...

This study develops a two-step estimator for a panel data Tobit model based on taking first-differences of the equation of interest, under conditional mean independence assumptions.The necessary correction terms are non-standard and a substantial part is therefore devoted to the formal...

In the linear model y = X + " under the restriction C = 0 a canonical partition C = [C0;C1] of the rows of C admits a simple representation of linear subspaces of values = X . Its use is shown for the identification, estimating and testing of linear combinations DB. Results are derived without...

In designing a production model for firms that generate multiple outputs, we take as a starting point that such multi-output production refers to economies of scope, which in turn originate from joint input use and input externalities. We provide a nonparametric characterization of cost...

We introduce generalized Probability-Probability (P-P) plots in order to study the one-sample goodness-of-fit problem and the two-sample problem, for real valued data.These plots, that are constructed by indexing with the class of closed intervals, globally preserve the properties of classical...

Integer-valued autoregressive (INAR) processes have been introduced to model nonnegative integervalued phenomena that evolve in time.The distribution of an INAR(p) process is determined by two parameters: a vector of survival probabilities and a probability distribution on the nonnegative...

In this paper we present a framework for backtesting all currently popular risk measurement methods (including value-at-risk and expected shortfall) using the functional delta method.Estimation risk can be taken explicitly into account.Based on a simulation study we provide evidence that tests...

High breakdown-point regression estimators protect against large errors and data contamination. We adapt and generalize the concept of trimming used by many of these robust estimators so that it can be employed in the context of the generalized method of moments. The proposed generalized method...