Marshall-Olkin distributions and portfolio credit risk
|
|
|
- Branden Pope
- 10 years ago
- Views:
Transcription
1 Marshall-Olkin distributions and portfolio credit risk Moderne Finanzmathematik und ihre Anwendungen für Banken und Versicherungen, Fraunhofer ITWM, Kaiserslautern, in Kooperation mit der TU München und DEVnet GmbH & Co KG, December 4, 2009 Jan-Frederik Mai HVB-Institute for Mathematical Finance Technische Universität München Joint work with Matthias Scherer.
2 Overview 1. Motivation 2. Exchangeable Marshall-Olkin distributions 3. The Lévy-frailty default model 4. Conclusion
3 Motivation: CDO pricing Situation: > Consider a portfolio of d credit-risky assets.
4 Motivation: CDO pricing Situation: > Consider a portfolio of d credit-risky assets. > (τ 1,..., τ d ) vector of (random) default times of the underlyings.
5 Motivation: CDO pricing Situation: > Consider a portfolio of d credit-risky assets. > (τ 1,..., τ d ) vector of (random) default times of the underlyings. > Typically, d = 125, i.e. large.
6 Motivation: CDO pricing Situation: > Consider a portfolio of d credit-risky assets. > (τ 1,..., τ d ) vector of (random) default times of the underlyings. > Typically, d = 125, i.e. large. > Constant and identical recovery rates, and equally weighted portfolio.
7 Motivation: CDO pricing Situation: > Consider a portfolio of d credit-risky assets. > (τ 1,..., τ d ) vector of (random) default times of the underlyings. > Typically, d = 125, i.e. large. > Constant and identical recovery rates, and equally weighted portfolio. > L t := 1 d d k=1 1 {τ k <t}= percentage of defaults up to time t 0.
8 Motivation: CDO pricing Situation: > Consider a portfolio of d credit-risky assets. > (τ 1,..., τ d ) vector of (random) default times of the underlyings. > Typically, d = 125, i.e. large. > Constant and identical recovery rates, and equally weighted portfolio. > L t := 1 d d k=1 1 {τ k <t}= percentage of defaults up to time t 0. CDO = financial contract whose payment streams depend on L t.
9 Motivation: CDO pricing Pricing of a CDO requires one to compute expectations of the form E [ f(l t ) ] = f(x) P(L t dx), f complicated. [0,1]
10 Motivation: CDO pricing Pricing of a CDO requires one to compute expectations of the form E [ f(l t ) ] = f(x) P(L t dx), f complicated. [0,1] Goal: Model (τ 1,..., τ d ) such that... (1)... one can approximate P(L t dx) efficiently (without Monte Carlo). (2)... dependence between the τ k is (economically) intuitive.
11 Latent One-Factor Copula Models Large Homogeneous Portfolio Assumption: > Infinite portfolio size, i.e. (τ 1,..., τ d ) {τ k } k N. > {τ k } k N is i.i.d. conditioned on a common market factor M.
12 Latent One-Factor Copula Models Large Homogeneous Portfolio Assumption: > Infinite portfolio size, i.e. (τ 1,..., τ d ) {τ k } k N. > {τ k } k N is i.i.d. conditioned on a common market factor M. Advantage: P(L t dx) P(M dx).
13 Latent One-Factor Copula Models Large Homogeneous Portfolio Assumption: > Infinite portfolio size, i.e. (τ 1,..., τ d ) {τ k } k N. > {τ k } k N is i.i.d. conditioned on a common market factor M. Advantage: P(L t dx) P(M dx). Examples: > [Li 2000]: M normal r.v. (τ 1,..., τ d ) Gaussian copula. > [Albrecher et al. 2007]: M ID r.v. (τ 1,..., τ d )???-copula. > [Schönbucher 2002]: M positive r.v. (τ 1,..., τ d ) Archimedean copula.
14 Latent One-Factor Copula Models Large Homogeneous Portfolio Assumption: > Infinite portfolio size, i.e. (τ 1,..., τ d ) {τ k } k N. > {τ k } k N is i.i.d. conditioned on a common market factor M. Advantage: P(L t dx) P(M dx). Examples: > [Li 2000]: M normal r.v. (τ 1,..., τ d ) Gaussian copula. > [Albrecher et al. 2007]: M ID r.v. (τ 1,..., τ d )???-copula. > [Schönbucher 2002]: M positive r.v. (τ 1,..., τ d ) Archimedean copula. Problem: (economically) intuitive?
15 Our idea: Marshall-Olkin copula Assume (τ 1,..., τ d ) Marshall-Olkin copula. Proposed by [Giesecke 2003; Embrechts et al. 2001; Lindskog, McNeil 2003].
16 Our idea: Marshall-Olkin copula Assume (τ 1,..., τ d ) Marshall-Olkin copula. Proposed by [Giesecke 2003; Embrechts et al. 2001; Lindskog, McNeil 2003]. Advantage: intuitive shock interpretation.
17 Our idea: Marshall-Olkin copula Assume (τ 1,..., τ d ) Marshall-Olkin copula. Proposed by [Giesecke 2003; Embrechts et al. 2001; Lindskog, McNeil 2003]. Advantage: intuitive shock interpretation. Problem: P(L t dx) intractable, since... > No latent market factor no large homogeneous portfolio assumption. > The model for (τ 1,..., τ d ) cannot be extended to {τ k } k N.
18 Our idea: Marshall-Olkin copula Assume (τ 1,..., τ d ) Marshall-Olkin copula. Proposed by [Giesecke 2003; Embrechts et al. 2001; Lindskog, McNeil 2003]. Advantage: intuitive shock interpretation. Problem: P(L t dx) intractable, since... > No latent market factor no large homogeneous portfolio assumption. > The model for (τ 1,..., τ d ) cannot be extended to {τ k } k N. Solution: We show how to define {τ k } k N appropriately.
19 Overview 1. Motivation 2. Exchangeable Marshall-Olkin distributions 3. The Lévy-frailty default model 4. Conclusion
20 Exchangeable Marshall-Olkin distributions Definition (in dimension 3) On a probability space (Ω, F, P) define the following independent random variables (= arrival times of exogeneous shocks ): E {1} E {2} E {3} Exp(λ 1 ) E {1,2} E {1,3} E {2,3} Exp(λ 2 ) E {1,2,3} Exp(λ 3 )
21 Exchangeable Marshall-Olkin distributions Definition (in dimension 3) On a probability space (Ω, F, P) define the following independent random variables (= arrival times of exogeneous shocks ): E {1} E {2} E {3} Exp(λ 1 ) E {1,2} E {1,3} E {2,3} Exp(λ 2 ) E {1,2,3} Exp(λ 3 ) Define the vector of extinction times ( τ 1, τ 2, τ 3) via τ 1 : = min { E {1} } { E {1,2}, E {1,3} } { E {1,2,3} } τ 2 : = min { E {2} } { E {1,2}, E {2,3} } { E{1,2,3} } τ 3 : = min { E {3} } { E{1,3}, E {2,3} } { E{1,2,3} } Exp(λ λ 2 + λ 3 ) Exp(λ λ 2 + λ 3 ) Exp(λ λ 2 + λ 3 )
22 Exchangeable Marshall-Olkin distributions General case Method of construction: [Marshall, Olkin (1967)] > Given: parameters λ 1,..., λ d > 0. > Consider (Ω, F, P) where E I Exp(λ I ) are independent, I {1,..., d}. > (τ 1,..., τ d ) has a so-called Marshall-Olkin distribution, where { } τ k := min EI, k = 1,..., d. I:k I
23 Exchangeable Marshall-Olkin distributions General case Method of construction: [Marshall, Olkin (1967)] > Given: parameters λ 1,..., λ d > 0. > Consider (Ω, F, P) where E I Exp(λ I ) are independent, I {1,..., d}. > (τ 1,..., τ d ) has a so-called Marshall-Olkin distribution, where { } τ k := min EI, k = 1,..., d. I:k I Intuitive Shock interpretation: E I = arrival time of an exogenous economy shock, destroys all components k I.
24 Exchangeable Marshall-Olkin copula Theorem: [Marshall, Olkin 1967] The exch. Marshall-Olkin distribution with parameters λ 1,..., λ d > 0 is given by ) ( ) P(τ 1 > t 1,..., τ d > t d ) = C d (P(τ 1 > t 1 ),..., P(τ d > t d ) = C d e O t 1,..., e O t d, where O := d 1 i=0 ( d 1 ) i λi+1 and C d (u 1,..., u d ) := d k=1 u 1 O d k i=0 ( d k i )λ i+1 (k), 0 u (1)... u (d) 1.
25 Exchangeable Marshall-Olkin copula Theorem: [Marshall, Olkin 1967] The exch. Marshall-Olkin distribution with parameters λ 1,..., λ d > 0 is given by ) ( ) P(τ 1 > t 1,..., τ d > t d ) = C d (P(τ 1 > t 1 ),..., P(τ d > t d ) = C d e O t 1,..., e O t d, where O := d 1 i=0 ( d 1 ) i λi+1 and C d (u 1,..., u d ) := d k=1 u 1 O d k i=0 ( d k i )λ i+1 (k), 0 u (1)... u (d) 1. C d is the survival copula of the exchangeable Marshall-Olkin distribution.
26 Suitable for portfolio credit risk? Problems: > P(L t dx) is complicated, especially for d 2.
27 Suitable for portfolio credit risk? Problems: > P(L t dx) is complicated, especially for d 2. > For large d 2, even Monte Carlo is impossible (O(2 d )).
28 Suitable for portfolio credit risk? Problems: > P(L t dx) is complicated, especially for d 2. > For large d 2, even Monte Carlo is impossible (O(2 d )). > Multiple shock model latent factor representation?
29 Suitable for portfolio credit risk? Problems: > P(L t dx) is complicated, especially for d 2. > For large d 2, even Monte Carlo is impossible (O(2 d )). > Multiple shock model latent factor representation? > We do not want exponential margins.
30 Suitable for portfolio credit risk? Problems: > P(L t dx) is complicated, especially for d 2. > For large d 2, even Monte Carlo is impossible (O(2 d )). > Multiple shock model latent factor representation? > We do not want exponential margins. Our contribution: We identify a very tractable subclass, where everything works out and interesting mathematical byproducts are obtained.
31 Overview 1. Motivation 2. Exchangeable Marshall-Olkin distributions 3. The Lévy-frailty default model 4. Conclusion
32 Lévy Subordinators (1) A Lévy subordinator Λ = {Λ t } t 0 is a non-decreasing Lévy process, i.e. > Λ 0 = 0 and non-decreasing > independent and stationary increments > stochastically continuous
33 Lévy Subordinators (1) A Lévy subordinator Λ = {Λ t } t 0 is a non-decreasing Lévy process, i.e. > Λ 0 = 0 and non-decreasing > independent and stationary increments > stochastically continuous Theorem: [Lévy 1937, Khinchin 1938] Λ Lévy subordinator Ψ : [0, ) [0, ) : E[e x Λ t ] = e t Ψ(x), x 0, t 0.
34 Lévy Subordinators (1) A Lévy subordinator Λ = {Λ t } t 0 is a non-decreasing Lévy process, i.e. > Λ 0 = 0 and non-decreasing > independent and stationary increments > stochastically continuous Theorem: [Lévy 1937, Khinchin 1938] Λ Lévy subordinator Ψ : [0, ) [0, ) : E[e x Λ t ] = e t Ψ(x), x 0, t 0. Ψ is called Laplace exponent, concave, in C (0, ).
35 Lévy Subordinators (2) Example 1:(Compound Poisson subordinator) Λ t = µ t + N t i=1 Ψ(x) = µ x + E[N 1 ] ( 1 E[e x J 1 ] ) J i, N = {N t } t 0 Poisson process, {J i } i N i.i.d. > 0, Λ(t) (Compound Poisson with Drift) Λ(t) (Compound Poisson with Drift) Time t Time t
36 Lévy Subordinators (3) Example 2:(Inverse Gaussian subordinator) Λ t = inf{s > 0 : η s + W s = β t}, Ψ(x) = β ( 2 x + η2 η ) {W s } s 0 Brownian motion, Λ(t) (Inverse Gaussian) Λ(t) (Inverse Gaussian) Time t Time t
37 Lévy Subordinators (4) Example 3:(Stable subordinator) For α [0, 1] there exists a Lévy subordinator Λ with Ψ(x) = x 1 α Λ(t) (α stable) Λ(t) (α stable) Time t Time t
38 Lévy-frailty construction Main Theorem: > G = continuous distribution function,, G(0) = 0, G(t) < 1 for all t > 0.
39 Lévy-frailty construction Main Theorem: > G = continuous distribution function,, G(0) = 0, G(t) < 1 for all t > 0. > Λ = {Λ t } t 0 Lévy subordinator satisfying Ψ(1) = 1.
40 Lévy-frailty construction Main Theorem: > G = continuous distribution function,, G(0) = 0, G(t) < 1 for all t > 0. > Λ = {Λ t } t 0 Lévy subordinator satisfying Ψ(1) = 1. > {E k } k N i.i.d. with E 1 Exp(1), independent of Λ.
41 Lévy-frailty construction Main Theorem: > G = continuous distribution function,, G(0) = 0, G(t) < 1 for all t > 0. > Λ = {Λ t } t 0 Lévy subordinator satisfying Ψ(1) = 1. > {E k } k N i.i.d. with E 1 Exp(1), independent of Λ. Define: τ k := inf { t > 0 : Λ log(1 G(t)) > E k }, k N.
42 Lévy-frailty construction Main Theorem: > G = continuous distribution function,, G(0) = 0, G(t) < 1 for all t > 0. > Λ = {Λ t } t 0 Lévy subordinator satisfying Ψ(1) = 1. > {E k } k N i.i.d. with E 1 Exp(1), independent of Λ. Define: τ k := inf { t > 0 : Λ log(1 G(t)) > E k }, k N. Then: P(τ 1 > t 1,..., τ d > t d ) = C d ( 1 G(t1 ),..., 1 G(t d ) ), where C d is the survival copula of a certain Marshall-Olkin distribution.
43 Some Remarks G(t) = P(τ k t)= default probability at time t 0 (known beforehand).
44 Some Remarks G(t) = P(τ k t)= default probability at time t 0 (known beforehand). Above theorem constitutes a (proper) subclass of exch. Marshall-Olkin copulas.
45 Some Remarks G(t) = P(τ k t)= default probability at time t 0 (known beforehand). Above theorem constitutes a (proper) subclass of exch. Marshall-Olkin copulas. Is this alternative construction useful?
46 Some Remarks G(t) = P(τ k t)= default probability at time t 0 (known beforehand). Above theorem constitutes a (proper) subclass of exch. Marshall-Olkin copulas. Is this alternative construction useful? Yes. Now we have a latent one-factor model: > G := σ(λ t : t 0) path of a Lévy subordinator. > (τ 1,..., τ d ) are i.i.d. conditioned on the latent factor G.
47 The Portfolio Loss Distribution Exact portfolio loss distribution: ( ) d m ( ) m P(L t = m) = ( 1) k (1 G(t)) Ψ(d+k m) m k k=0 Example: Ψ(x) = x 1 α, G(t) 26%, d = α= α= α=
48 Approximation of Portfolio Loss Corollary: As the portfolio size d, it holds that {L t } t [0,T ] L 2 { } 1 e Λ log(1 G(t)) t [0,T ].
49 Approximation of Portfolio Loss Corollary: As the portfolio size d, it holds that {L t } t [0,T ] L 2 In particular, provided Λ t has a density f Λt { } 1 e Λ log(1 G(t)) t [0,T ]. for all t > 0, we have that ( ) 1 f Lt (x) f Λ log(1 G(t)) log(1 x), x (0, 1). 1 x
50 Approximation of Portfolio Loss Corollary: As the portfolio size d, it holds that {L t } t [0,T ] L 2 In particular, provided Λ t has a density f Λt { } 1 e Λ log(1 G(t)) t [0,T ]. for all t > 0, we have that ( ) 1 f Lt (x) f Λ log(1 G(t)) log(1 x), x (0, 1). 1 x Thus, we can efficiently compute E [ f(l t ) ] = f(x) P(L t dx) [0,1] [0,1] f(x) f Λ log(1 G(t)) ( log(1 x) ) 1 1 x dx.
51 Inverse Gaussian density of the portfolio loss One-parametric Inverse Gaussian Lévy subordinator (parameter η > 0): ) ( ( f Lt (x) log(1 G(t)) η log(1 G(t)) log(1 G(t)) log 2 (1 x) 2 π ( 2+η 2 η) exp ( 2+η 2 η x) exp 2 ( log(1 x)) ( 2+η η2 log(1 x) 2 η) 2 2 ), x 0. density value η = 10 η = 1 η = loss percentage
52 Overview 1. Motivation 2. Exchangeable Marshall-Olkin distributions 3. The Lévy-frailty default model 4. Conclusion
53 Conclusion Latent factor representation for certain Marshall-Olkin distributions
54 Conclusion Latent factor representation for certain Marshall-Olkin distributions Very useful to study Marshall-Olkin distributions
55 Conclusion Latent factor representation for certain Marshall-Olkin distributions Very useful to study Marshall-Olkin distributions Interesting coherences between copulas, completely monotone sequences, Lévy subordinators
56 Conclusion Latent factor representation for certain Marshall-Olkin distributions Very useful to study Marshall-Olkin distributions Interesting coherences between copulas, completely monotone sequences, Lévy subordinators Efficient approximation of portfolio loss distribution quick calibration to market data
57 Conclusion Latent factor representation for certain Marshall-Olkin distributions Very useful to study Marshall-Olkin distributions Interesting coherences between copulas, completely monotone sequences, Lévy subordinators Efficient approximation of portfolio loss distribution quick calibration to market data The model is more dynamic than state-of-the-art one-factor models, since latent factor is a stochastic process
58 References with M. Scherer: A tractable multivariate default model based on a stochastic time-change. International Journal of Theoretical and Applied Finance, 12:2, pp (2009). with M. Scherer: Lévy-Frailty Copulas. Journal of Multivariate Analysis, 100:7, pp (2009). with M. Scherer: Reparameterizing Marshall-Olkin copulas with applications to sampling. Journal of Statistical Computation and Simulation (in press) (2009).
59 Thank you for your attention.
Credit Risk Models: An Overview
Credit Risk Models: An Overview Paul Embrechts, Rüdiger Frey, Alexander McNeil ETH Zürich c 2003 (Embrechts, Frey, McNeil) A. Multivariate Models for Portfolio Credit Risk 1. Modelling Dependent Defaults:
Pricing multi-asset options and credit derivatives with copulas
Pricing multi-asset options and credit derivatives with copulas Séminaire de Mathématiques et Finance Louis Bachelier, Institut Henri Poincaré Thierry Roncalli Groupe de Recherche Opérationnelle Crédit
Pricing of a worst of option using a Copula method M AXIME MALGRAT
Pricing of a worst of option using a Copula method M AXIME MALGRAT Master of Science Thesis Stockholm, Sweden 2013 Pricing of a worst of option using a Copula method MAXIME MALGRAT Degree Project in Mathematical
Point Process Techniques in Non-Life Insurance Models
1 Point Process Techniques in Non-Life Insurance Models Thomas Mikosch University of Copenhagen 1 Conference in Honor of Jan Grandell, Stockholm, June 13, 2008 1 2 What is a point process? 3 Consider the
Stephane Crepey. Financial Modeling. A Backward Stochastic Differential Equations Perspective. 4y Springer
Stephane Crepey Financial Modeling A Backward Stochastic Differential Equations Perspective 4y Springer Part I An Introductory Course in Stochastic Processes 1 Some Classes of Discrete-Time Stochastic
Monte Carlo Simulation
1 Monte Carlo Simulation Stefan Weber Leibniz Universität Hannover email: [email protected] web: www.stochastik.uni-hannover.de/ sweber Monte Carlo Simulation 2 Quantifying and Hedging
Math 526: Brownian Motion Notes
Math 526: Brownian Motion Notes Definition. Mike Ludkovski, 27, all rights reserved. A stochastic process (X t ) is called Brownian motion if:. The map t X t (ω) is continuous for every ω. 2. (X t X t
Portfolio Distribution Modelling and Computation. Harry Zheng Department of Mathematics Imperial College [email protected]
Portfolio Distribution Modelling and Computation Harry Zheng Department of Mathematics Imperial College [email protected] Workshop on Fast Financial Algorithms Tanaka Business School Imperial College
A spot price model feasible for electricity forward pricing Part II
A spot price model feasible for electricity forward pricing Part II Fred Espen Benth Centre of Mathematics for Applications (CMA) University of Oslo, Norway Wolfgang Pauli Institute, Wien January 17-18
Monte Carlo Methods in Finance
Author: Yiyang Yang Advisor: Pr. Xiaolin Li, Pr. Zari Rachev Department of Applied Mathematics and Statistics State University of New York at Stony Brook October 2, 2012 Outline Introduction 1 Introduction
Overview of Monte Carlo Simulation, Probability Review and Introduction to Matlab
Monte Carlo Simulation: IEOR E4703 Fall 2004 c 2004 by Martin Haugh Overview of Monte Carlo Simulation, Probability Review and Introduction to Matlab 1 Overview of Monte Carlo Simulation 1.1 Why use simulation?
Valuation of a Homogeneous Collateralized Debt Obligation
Valuation of a Homogeneous Collateralized Debt Obligation by Fabio Mibielli Peixoto 1 An essay presented to the University of Waterloo in partial fulfillment of the requirements for the degree of Masters
Monte Carlo and Empirical Methods for Stochastic Inference (MASM11/FMS091)
Monte Carlo and Empirical Methods for Stochastic Inference (MASM11/FMS091) Magnus Wiktorsson Centre for Mathematical Sciences Lund University, Sweden Lecture 5 Sequential Monte Carlo methods I February
1 Simulating Brownian motion (BM) and geometric Brownian motion (GBM)
Copyright c 2013 by Karl Sigman 1 Simulating Brownian motion (BM) and geometric Brownian motion (GBM) For an introduction to how one can construct BM, see the Appendix at the end of these notes A stochastic
arxiv:1412.1183v1 [q-fin.rm] 3 Dec 2014
Regulatory Capital Modelling for Credit Risk arxiv:1412.1183v1 [q-fin.rm] 3 Dec 2014 Marek Rutkowski a, Silvio Tarca a, a School of Mathematics and Statistics F07, University of Sydney, NSW 2006, Australia.
Monte Carlo-based statistical methods (MASM11/FMS091)
Monte Carlo-based statistical methods (MASM11/FMS091) Jimmy Olsson Centre for Mathematical Sciences Lund University, Sweden Lecture 5 Sequential Monte Carlo methods I February 5, 2013 J. Olsson Monte Carlo-based
Bias in the Estimation of Mean Reversion in Continuous-Time Lévy Processes
Bias in the Estimation of Mean Reversion in Continuous-Time Lévy Processes Yong Bao a, Aman Ullah b, Yun Wang c, and Jun Yu d a Purdue University, IN, USA b University of California, Riverside, CA, USA
A Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails
12th International Congress on Insurance: Mathematics and Economics July 16-18, 2008 A Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails XUEMIAO HAO (Based on a joint
LÉVY-DRIVEN PROCESSES IN BAYESIAN NONPARAMETRIC INFERENCE
Bol. Soc. Mat. Mexicana (3) Vol. 19, 2013 LÉVY-DRIVEN PROCESSES IN BAYESIAN NONPARAMETRIC INFERENCE LUIS E. NIETO-BARAJAS ABSTRACT. In this article we highlight the important role that Lévy processes have
On exponentially ane martingales. Johannes Muhle-Karbe
On exponentially ane martingales AMAMEF 2007, Bedlewo Johannes Muhle-Karbe Joint work with Jan Kallsen HVB-Institut für Finanzmathematik, Technische Universität München 1 Outline 1. Semimartingale characteristics
Grey Brownian motion and local times
Grey Brownian motion and local times José Luís da Silva 1,2 (Joint work with: M. Erraoui 3 ) 2 CCM - Centro de Ciências Matemáticas, University of Madeira, Portugal 3 University Cadi Ayyad, Faculty of
From Ruin Theory to Solvency in Non-Life Insurance
From Ruin Theory to Solvency in Non-Life Insurance Mario V. Wüthrich RiskLab ETH Zurich & Swiss Finance Institute SFI January 23, 2014 LUH Colloquium Versicherungs- und Finanzmathematik Leibniz Universität
Corrected Diffusion Approximations for the Maximum of Heavy-Tailed Random Walk
Corrected Diffusion Approximations for the Maximum of Heavy-Tailed Random Walk Jose Blanchet and Peter Glynn December, 2003. Let (X n : n 1) be a sequence of independent and identically distributed random
A LOGNORMAL MODEL FOR INSURANCE CLAIMS DATA
REVSTAT Statistical Journal Volume 4, Number 2, June 2006, 131 142 A LOGNORMAL MODEL FOR INSURANCE CLAIMS DATA Authors: Daiane Aparecida Zuanetti Departamento de Estatística, Universidade Federal de São
FORWARDS AND EUROPEAN OPTIONS ON CDO TRANCHES. John Hull and Alan White. First Draft: December, 2006 This Draft: March 2007
FORWARDS AND EUROPEAN OPTIONS ON CDO TRANCHES John Hull and Alan White First Draft: December, 006 This Draft: March 007 Joseph L. Rotman School of Management University of Toronto 105 St George Street
Stocks paying discrete dividends: modelling and option pricing
Stocks paying discrete dividends: modelling and option pricing Ralf Korn 1 and L. C. G. Rogers 2 Abstract In the Black-Scholes model, any dividends on stocks are paid continuously, but in reality dividends
Modelling cycle dependence in credit insurance
Modelling cycle dependence in credit insurance - Anisa CAJA (Université Lyon 1, Laboratoire SAF) - Frédéric PLANCHET (Université Lyon 1, Laboratoire SAF) 2013.18 Laboratoire SAF 50 Avenue Tony Garnier
Credit Derivatives: fundaments and ideas
Credit Derivatives: fundaments and ideas Roberto Baviera, Rates & Derivatives Trader & Structurer, Abaxbank I.C.T.P., 14-17 Dec 2007 1 About? 2 lacked knowledge Source: WSJ 5 Dec 07 3 Outline 1. overview
Monte Carlo Methods and Models in Finance and Insurance
Chapman & Hall/CRC FINANCIAL MATHEMATICS SERIES Monte Carlo Methods and Models in Finance and Insurance Ralf Korn Elke Korn Gerald Kroisandt f r oc) CRC Press \ V^ J Taylor & Francis Croup ^^"^ Boca Raton
Pricing and calibration in local volatility models via fast quantization
Pricing and calibration in local volatility models via fast quantization Parma, 29 th January 2015. Joint work with Giorgia Callegaro and Martino Grasselli Quantization: a brief history Birth: back to
Asymptotics for ruin probabilities in a discrete-time risk model with dependent financial and insurance risks
1 Asymptotics for ruin probabilities in a discrete-time risk model with dependent financial and insurance risks Yang Yang School of Mathematics and Statistics, Nanjing Audit University School of Economics
Maximum Likelihood Estimation
Math 541: Statistical Theory II Lecturer: Songfeng Zheng Maximum Likelihood Estimation 1 Maximum Likelihood Estimation Maximum likelihood is a relatively simple method of constructing an estimator for
Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations
56 Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations Florin-Cătălin ENACHE
Exponential Distribution
Exponential Distribution Definition: Exponential distribution with parameter λ: { λe λx x 0 f(x) = 0 x < 0 The cdf: F(x) = x Mean E(X) = 1/λ. f(x)dx = Moment generating function: φ(t) = E[e tx ] = { 1
Statistics Graduate Courses
Statistics Graduate Courses STAT 7002--Topics in Statistics-Biological/Physical/Mathematics (cr.arr.).organized study of selected topics. Subjects and earnable credit may vary from semester to semester.
Non-Life Insurance Mathematics
Thomas Mikosch Non-Life Insurance Mathematics An Introduction with the Poisson Process Second Edition 4y Springer Contents Part I Collective Risk Models 1 The Basic Model 3 2 Models for the Claim Number
How to Model Operational Risk, if You Must
How to Model Operational Risk, if You Must Paul Embrechts ETH Zürich (www.math.ethz.ch/ embrechts) Based on joint work with V. Chavez-Demoulin, H. Furrer, R. Kaufmann, J. Nešlehová and G. Samorodnitsky
Two Topics in Parametric Integration Applied to Stochastic Simulation in Industrial Engineering
Two Topics in Parametric Integration Applied to Stochastic Simulation in Industrial Engineering Department of Industrial Engineering and Management Sciences Northwestern University September 15th, 2014
EXIT TIME PROBLEMS AND ESCAPE FROM A POTENTIAL WELL
EXIT TIME PROBLEMS AND ESCAPE FROM A POTENTIAL WELL Exit Time problems and Escape from a Potential Well Escape From a Potential Well There are many systems in physics, chemistry and biology that exist
Master s Theory Exam Spring 2006
Spring 2006 This exam contains 7 questions. You should attempt them all. Each question is divided into parts to help lead you through the material. You should attempt to complete as much of each problem
Contents. List of Figures. List of Tables. List of Examples. Preface to Volume IV
Contents List of Figures List of Tables List of Examples Foreword Preface to Volume IV xiii xvi xxi xxv xxix IV.1 Value at Risk and Other Risk Metrics 1 IV.1.1 Introduction 1 IV.1.2 An Overview of Market
Estimating the Degree of Activity of jumps in High Frequency Financial Data. joint with Yacine Aït-Sahalia
Estimating the Degree of Activity of jumps in High Frequency Financial Data joint with Yacine Aït-Sahalia Aim and setting An underlying process X = (X t ) t 0, observed at equally spaced discrete times
VISUALIZATION OF DENSITY FUNCTIONS WITH GEOGEBRA
VISUALIZATION OF DENSITY FUNCTIONS WITH GEOGEBRA Csilla Csendes University of Miskolc, Hungary Department of Applied Mathematics ICAM 2010 Probability density functions A random variable X has density
A SURVEY ON DEFAULT TIMES AS A TOOL FOR PORTFOLIO MANAGEMENT
A SURVEY ON DEFAULT TIMES AS A TOOL FOR PORTFOLIO MANAGEMENT RISK DAY, ETH ZURICH, OCTOBER 15, 2004 Dr. Christian Bluhm Head Credit Suisse, Zurich [email protected] AGENDA INTRODUCTION
INSURANCE RISK THEORY (Problems)
INSURANCE RISK THEORY (Problems) 1 Counting random variables 1. (Lack of memory property) Let X be a geometric distributed random variable with parameter p (, 1), (X Ge (p)). Show that for all n, m =,
Spatial Statistics Chapter 3 Basics of areal data and areal data modeling
Spatial Statistics Chapter 3 Basics of areal data and areal data modeling Recall areal data also known as lattice data are data Y (s), s D where D is a discrete index set. This usually corresponds to data
Concentration inequalities for order statistics Using the entropy method and Rényi s representation
Concentration inequalities for order statistics Using the entropy method and Rényi s representation Maud Thomas 1 in collaboration with Stéphane Boucheron 1 1 LPMA Université Paris-Diderot High Dimensional
Some remarks on two-asset options pricing and stochastic dependence of asset prices
Some remarks on two-asset options pricing and stochastic dependence of asset prices G. Rapuch & T. Roncalli Groupe de Recherche Opérationnelle, Crédit Lyonnais, France July 16, 001 Abstract In this short
Chapter 1. Introduction
Chapter 1 Introduction 1.1. Motivation Network performance analysis, and the underlying queueing theory, was born at the beginning of the 20th Century when two Scandinavian engineers, Erlang 1 and Engset
Pricing and Risk Management of Variable Annuity Guaranteed Benefits by Analytical Methods Longevity 10, September 3, 2014
Pricing and Risk Management of Variable Annuity Guaranteed Benefits by Analytical Methods Longevity 1, September 3, 214 Runhuan Feng, University of Illinois at Urbana-Champaign Joint work with Hans W.
Fast Monte Carlo CVA using Exposure Sampling Method
Fast Monte Carlo CVA using Exposure Sampling Method Alexander Sokol Numerix RiskMinds Conference 2010 (Geneva) Definitions Potential Future Exposure (PFE) PFE(T) is maximum loss due to counterparty default
Probability and Random Variables. Generation of random variables (r.v.)
Probability and Random Variables Method for generating random variables with a specified probability distribution function. Gaussian And Markov Processes Characterization of Stationary Random Process Linearly
The Exponential Distribution
21 The Exponential Distribution From Discrete-Time to Continuous-Time: In Chapter 6 of the text we will be considering Markov processes in continuous time. In a sense, we already have a very good understanding
Basics of Statistical Machine Learning
CS761 Spring 2013 Advanced Machine Learning Basics of Statistical Machine Learning Lecturer: Xiaojin Zhu [email protected] Modern machine learning is rooted in statistics. You will find many familiar
Introduction to Markov Chain Monte Carlo
Introduction to Markov Chain Monte Carlo Monte Carlo: sample from a distribution to estimate the distribution to compute max, mean Markov Chain Monte Carlo: sampling using local information Generic problem
BayesX - Software for Bayesian Inference in Structured Additive Regression
BayesX - Software for Bayesian Inference in Structured Additive Regression Thomas Kneib Faculty of Mathematics and Economics, University of Ulm Department of Statistics, Ludwig-Maximilians-University Munich
Lecture 13: Martingales
Lecture 13: Martingales 1. Definition of a Martingale 1.1 Filtrations 1.2 Definition of a martingale and its basic properties 1.3 Sums of independent random variables and related models 1.4 Products of
M/M/1 and M/M/m Queueing Systems
M/M/ and M/M/m Queueing Systems M. Veeraraghavan; March 20, 2004. Preliminaries. Kendall s notation: G/G/n/k queue G: General - can be any distribution. First letter: Arrival process; M: memoryless - exponential
Options On Credit Default Index Swaps
Options On Credit Default Index Swaps Yunkang Liu and Peter Jäckel 20th May 2005 Abstract The value of an option on a credit default index swap consists of two parts. The first one is the protection value
Quantitative Operational Risk Management
Quantitative Operational Risk Management Kaj Nyström and Jimmy Skoglund Swedbank, Group Financial Risk Control S-105 34 Stockholm, Sweden September 3, 2002 Abstract The New Basel Capital Accord presents
COMPLETE MARKETS DO NOT ALLOW FREE CASH FLOW STREAMS
COMPLETE MARKETS DO NOT ALLOW FREE CASH FLOW STREAMS NICOLE BÄUERLE AND STEFANIE GRETHER Abstract. In this short note we prove a conjecture posed in Cui et al. 2012): Dynamic mean-variance problems in
Monte Carlo-based statistical methods (MASM11/FMS091)
Monte Carlo-based statistical methods (MASM11/FMS091) Magnus Wiktorsson Centre for Mathematical Sciences Lund University, Sweden Lecture 6 Sequential Monte Carlo methods II February 7, 2014 M. Wiktorsson
The Monte Carlo Framework, Examples from Finance and Generating Correlated Random Variables
Monte Carlo Simulation: IEOR E4703 Fall 2004 c 2004 by Martin Haugh The Monte Carlo Framework, Examples from Finance and Generating Correlated Random Variables 1 The Monte Carlo Framework Suppose we wish
Mathematical Finance
Mathematical Finance Option Pricing under the Risk-Neutral Measure Cory Barnes Department of Mathematics University of Washington June 11, 2013 Outline 1 Probability Background 2 Black Scholes for European
A Model of Optimum Tariff in Vehicle Fleet Insurance
A Model of Optimum Tariff in Vehicle Fleet Insurance. Bouhetala and F.Belhia and R.Salmi Statistics and Probability Department Bp, 3, El-Alia, USTHB, Bab-Ezzouar, Alger Algeria. Summary: An approach about
On the mathematical theory of splitting and Russian roulette
On the mathematical theory of splitting and Russian roulette techniques St.Petersburg State University, Russia 1. Introduction Splitting is an universal and potentially very powerful technique for increasing
Risk Management and Portfolio Optimization for Volatile Markets
Risk Management and Portfolio Optimization for Volatile Markets Abstract Svetlozar T. Rachev Chief Scientist, FinAnalytica and Chair Professor of Statistics, Econometrics and Mathematical Finance, School
FULL LIST OF REFEREED JOURNAL PUBLICATIONS Qihe Tang
FULL LIST OF REFEREED JOURNAL PUBLICATIONS Qihe Tang 87. Li, J.; Tang, Q. Interplay of insurance and financial risks in a discrete-time model with strongly regular variation. Bernoulli 21 (2015), no. 3,
Package fastghquad. R topics documented: February 19, 2015
Package fastghquad February 19, 2015 Type Package Title Fast Rcpp implementation of Gauss-Hermite quadrature Version 0.2 Date 2014-08-13 Author Alexander W Blocker Maintainer Fast, numerically-stable Gauss-Hermite
Aggregate Loss Models
Aggregate Loss Models Chapter 9 Stat 477 - Loss Models Chapter 9 (Stat 477) Aggregate Loss Models Brian Hartman - BYU 1 / 22 Objectives Objectives Individual risk model Collective risk model Computing
Tests for exponentiality against the M and LM-classes of life distributions
Tests for exponentiality against the M and LM-classes of life distributions B. Klar Universität Karlsruhe Abstract This paper studies tests for exponentiality against the nonparametric classes M and LM
Staffing and Control of Instant Messaging Contact Centers
OPERATIONS RESEARCH Vol. 61, No. 2, March April 213, pp. 328 343 ISSN 3-364X (print) ISSN 1526-5463 (online) http://dx.doi.org/1.1287/opre.112.1151 213 INFORMS Staffing and Control of Instant Messaging
e.g. arrival of a customer to a service station or breakdown of a component in some system.
Poisson process Events occur at random instants of time at an average rate of λ events per second. e.g. arrival of a customer to a service station or breakdown of a component in some system. Let N(t) be
A SURVEY ON CONTINUOUS ELLIPTICAL VECTOR DISTRIBUTIONS
A SURVEY ON CONTINUOUS ELLIPTICAL VECTOR DISTRIBUTIONS Eusebio GÓMEZ, Miguel A. GÓMEZ-VILLEGAS and J. Miguel MARÍN Abstract In this paper it is taken up a revision and characterization of the class of
Copula Concepts in Financial Markets
Copula Concepts in Financial Markets Svetlozar T. Rachev, University of Karlsruhe, KIT & University of Santa Barbara & FinAnalytica* Michael Stein, University of Karlsruhe, KIT** Wei Sun, University of
Assignment 2: Option Pricing and the Black-Scholes formula The University of British Columbia Science One CS 2015-2016 Instructor: Michael Gelbart
Assignment 2: Option Pricing and the Black-Scholes formula The University of British Columbia Science One CS 2015-2016 Instructor: Michael Gelbart Overview Due Thursday, November 12th at 11:59pm Last updated
Binomial lattice model for stock prices
Copyright c 2007 by Karl Sigman Binomial lattice model for stock prices Here we model the price of a stock in discrete time by a Markov chain of the recursive form S n+ S n Y n+, n 0, where the {Y i }
THE CENTRAL LIMIT THEOREM TORONTO
THE CENTRAL LIMIT THEOREM DANIEL RÜDT UNIVERSITY OF TORONTO MARCH, 2010 Contents 1 Introduction 1 2 Mathematical Background 3 3 The Central Limit Theorem 4 4 Examples 4 4.1 Roulette......................................
Introduction to Probability
Introduction to Probability EE 179, Lecture 15, Handout #24 Probability theory gives a mathematical characterization for experiments with random outcomes. coin toss life of lightbulb binary data sequence
Operational Risk and Insurance: Quantitative and qualitative aspects. Silke Brandts a
Operational Risk and Insurance: Quantitative and qualitative aspects Silke Brandts a Preliminary version - do not quote - This version: April 30, 2004 Abstract This paper incorporates insurance contracts
An exact formula for default swaptions pricing in the SSRJD stochastic intensity model
An exact formula for default swaptions pricing in the SSRJD stochastic intensity model Naoufel El-Bachir (joint work with D. Brigo, Banca IMI) Radon Institute, Linz May 31, 2007 ICMA Centre, University
LogNormal stock-price models in Exams MFE/3 and C/4
Making sense of... LogNormal stock-price models in Exams MFE/3 and C/4 James W. Daniel Austin Actuarial Seminars http://www.actuarialseminars.com June 26, 2008 c Copyright 2007 by James W. Daniel; reproduction
STA 4273H: Statistical Machine Learning
STA 4273H: Statistical Machine Learning Russ Salakhutdinov Department of Statistics! [email protected]! http://www.cs.toronto.edu/~rsalakhu/ Lecture 6 Three Approaches to Classification Construct
Bayesian Statistics: Indian Buffet Process
Bayesian Statistics: Indian Buffet Process Ilker Yildirim Department of Brain and Cognitive Sciences University of Rochester Rochester, NY 14627 August 2012 Reference: Most of the material in this note
Probability Theory. Florian Herzog. A random variable is neither random nor variable. Gian-Carlo Rota, M.I.T..
Probability Theory A random variable is neither random nor variable. Gian-Carlo Rota, M.I.T.. Florian Herzog 2013 Probability space Probability space A probability space W is a unique triple W = {Ω, F,
Introduction to General and Generalized Linear Models
Introduction to General and Generalized Linear Models General Linear Models - part I Henrik Madsen Poul Thyregod Informatics and Mathematical Modelling Technical University of Denmark DK-2800 Kgs. Lyngby
Alessandro Birolini. ineerin. Theory and Practice. Fifth edition. With 140 Figures, 60 Tables, 120 Examples, and 50 Problems.
Alessandro Birolini Re ia i it En ineerin Theory and Practice Fifth edition With 140 Figures, 60 Tables, 120 Examples, and 50 Problems ~ Springer Contents 1 Basic Concepts, Quality and Reliability Assurance
APPLYING COPULA FUNCTION TO RISK MANAGEMENT. Claudio Romano *
APPLYING COPULA FUNCTION TO RISK MANAGEMENT Claudio Romano * Abstract This paper is part of the author s Ph. D. Thesis Extreme Value Theory and coherent risk measures: applications to risk management.
