Modern risk systems
Papers published in 2015

Companion sites: Russian version


  1. Y. Kopeliovich, A. Novosyolov, D. Satchkov, B. Schachter. Robust Risk Estimation and Hedging: A Reverse Stress Testing Approach, The Journal of Derivatives, Vol. 22 (2015), No. 4, 10-25.
    Abstract
    The stress test has become an increasingly important risk assessment and management tool. But while it is easy to imagine a stress scenario and to estimate its impact on the firms financial condition, it is not so obvious how to select the most meaningful scenarios in the first place, either to get reasonable coverage of the space of stressful possibilities or even to focus on those that are most probable. In this article, the authors approach the problem from the reverse direction. They begin with a specified level of loss and pick the most likely scenario that generates that loss. They then use principal components to construct a set of alternative scenarios that produce the same level of loss but in (maximally) different ways. This provides much greater insight into which sources of risk are the most important and the most stable across scenarios.


  2. A. Novosyolov. Risk Aversion Concept in Preference Models, Proceedings of the XIV Internaional Conference on Financial and Actuarial Mathematics and Eventology of Multidimensional Statistics, Krasnoyarsk, SFU, 2015, 313-317.
    Abstract
    We consider methods of measuring risk aversion in some prefeerence models. In particular, we establish links among risk aversion parameter of exponential expected utility model, required return in the Markowitz problem, and certainty equivalent. We also suggest an idea of measuring risk aversion in coherent risk measures model.
    Download

  3. A. Novosyolov. Monte Carlo method: Variants of Application, Risk management in a Credit Organization, 1 (2015), 107-110.
    Abstract
    The study note describes the basic Monte Carlo method idea, and presents some modes of the method application: approximate calculation of probabilities of complex events, computing distributions of complex combinations of random variables, approximate calculation of integrals.


  4. A. Novosyolov. Monte Carlo method: Generating Distributions, Risk management in a Credit Organization, 2 (2015), 108-112.
    Abstract
    The study note describes methods of generating univariate random variables with given probability distributions, which are based on inverse distribution function, applying known relations with other distributions, and a number of specific tricks.


  5. A. Novosyolov. Monte Carlo method: Generating Miltivariate Distributions, Risk management in a Credit Organization, 3 (2015), 102-106.
    Abstract
    The study note describes methods of generating random vectors from multivariate distributions, based on Choleski and similar trasnforms, and on the distribution sense. We also present a method of generating a uniform distribution over standard simplex in Rn>.


  6. A. Novosyolov. Monte Carlo method using copula, Risk management in a Credit Organization, 4 (2015), 102-105.
    Abstract
    The study note describes methods of generating multivariate probability distributions based on separate representation of deoendence (copula function) and univariate (marginal) distributions of the components.



Home Basics Lectures Papers Download Links Contact
Copyright © 2000-2017, A.Novosyolov Last changed at 15.05.2016