Modern risk systems
Papers published in 2006

Companion sites: Russian version


  1. Novosyolov A.A. The sum of dependent normal variables may be not normal. Working paper, 2006, 13p.

    Abstract. This brief note provides calculation of distributions in the example of dependent normal variables with non-normal sum. The example belongs to Glyn Holton and intended to break the illusion of normality of linear combinations of normal variables.

    Download
  2. Novosyolov A.A. Representing complete and partial preferences over sets of risks by real functionals. Proceedings of the V All-Russian FAM conference, v. 1, Krasnoyarsk, 2006. (translated from Russian)

    Abstract. The paper is devoted to approximation of a functional, representing a preference relation over a set of probability distributions in case of informational deficit. Representation theorems are established for complete and partial preferences. Partitions generated by preference relations have been studied; a completing procedure in terms of representing families has been proposed, and examples of applications have been provided.

    Unavailable for downloading.
  3. Novosyolov A.A. Certainty equivalent in models of decision-making under risk. Proceedings of the V All-Russian FAM conference, v. 1, Krasnoyarsk, 2006. (translated from Russian)

    Abstract. Certainty equivalents for classic risk measures and for the generalized coherent risk measure have been presented.

    Unavailable for downloading.
  4. Novosyolov A.A. On a sense of higher order stochastic dominance. Proceedings of the V All-Russian FAM conference, v. 1, Krasnoyarsk, 2006. (translated from Russian)

    Abstract. Integral and asymptotic representations for integral distribution functions have been obtained. Examples explaning the sense of direct and dual second order stochastic dominance in terms of insurance and options trading are also presented.

    Unavailable for downloading.
  5. Novosyolov A.A. Representation of preferences on a set of risks by functional families. Proceedings of the International Scientific School "Modelling and Analysis of Safety and Risk in Complex Systems", St.-Petersburg, 2006, p. 92-97.

    Abstract. Quantitative decision-making under risk often reduces to solving optimization problems. The reduction requires representation of decision-makerís preferences by real-valued functionals. Personal preferences over risky outcomes in practice are seldom known completely. Partial preference is a more reasonable object to be assumed known. The paper is devoted to representation of personal preferences over sets of probability distributions by real functionals and functional families, and to approximation of complete preferences by partial ones in terms of representing families of functionals.

    Download
    paper, presentation
  6. Novosyolov A.A. Copula function as a dependence structure. Proceedings of the International Scientific School "Modelling and Analysis of Safety and Risk in Complex Systems", St.-Petersburg, 2006, p. 98-99.

    Abstract. Correlation structure of a multidimensional probability distribution captures only a small fraction of information on components dependence. To overcome the lack of information copula functions were intensively used during recent decades. The present paper gives an overview of copula functions as descriptors of dependence structure, presents some general classes of copula functions, and provides algorithms for using copula functions in decision-making under risk. Special attention is paid to discrete marginals case, which has not been addressed much in the literature before.

    Download
    paper, presentation

Home Basics Lectures Papers Download Links Contact
Copyright © 2000-2017, A.Novosyolov Last changed at 22.07.2015