Composite Tukey-type distributions with application to operational risk management

Linda Möstel,Matthias Fischer,Marius Pfeuffer

DOI: https://doi.org/10.21314/jop.2023.010

IF: 0.645

2024-03-05

The Journal of Operational Risk

Abstract:This paper investigates composite Tukey-type distributions and puts forward a new composite model, the improved flexibility of which is demonstrated.

business, finance

What problem does this paper attempt to address?

Multivariate Risk Models under Heavy-Tailed Risks

Wei Huang,Chengguo Weng,Yi Zhang

DOI: https://doi.org/10.1002/asmb.1981

2014-01-01

Applied Stochastic Models in Business and Industry

Abstract:In this paper, we consider four common types of ruin probabilities for a discrete-time multivariate risk model, where the insurer is assumed to be exposed to a vector of net losses resulting from a number of business lines over each period. By assuming a large initial capital for the risk model and regularly varying distributions for the net losses, we establish some interesting asymptotic estimates for ruin probabilities in terms of the upper tail dependence function of the net loss vector. Our results insightfully characterize how the dependence structure among the individual net losses affect the ruin probabilities in an asymptotic sense, and more importantly, from our main results, explicit asymptotic estimates for those ruin probabilities can be obtained via specifying a copula for the net loss vectors. Copyright © 2013 John Wiley & Sons, Ltd.
Data-driven and distribution-free estimation of tailed-related risks for GARCH models using composite asymmetric least squares regression

Jun Zhao,Yi Zhang,Sheng Wu,Liming Shen

DOI: https://doi.org/10.1016/j.cam.2021.113862

IF: 2.872

2022-01-01

Journal of Computational and Applied Mathematics

Abstract:In this paper, we propose a two-step procedure to estimate tail-related risks like VaR and ES for the GARCH model. We inventively put forward the composite asymmetric least squares (CALS) regression to estimate the volatility structure and distinguish it from the innovation process in the GARCH model. Then noting that expectile bridges the gap between VaR and ES, we introduce the empirical likelihood method to determine these relations. Accordingly, a new optimization algorithm is proposed. Compared with the existing grid-search method, this new proposed method is data-driven and distribution-free, and shares better estimation accuracy and computational efficiency. Monte Carlo simulation studies and empirical analysis also indicate that our proposed method is superior to some alternative existing tail-related risk estimation methods.
Some analytical results on bivariate stable distributions with an application in operational risk

L. Tafakori,M. Bee,A.R. Soltani

DOI: https://doi.org/10.1080/14697688.2022.2046285

IF: 1.986

2022-03-24

Quantitative Finance

Abstract:The multivariate stable distributions are widely applicable as they can accommodate both skewness and heavy tails. Although one-dimensional stable distributions are well known, there are many open questions in the multivariate regime, since the tractability of the multivariate Gaussian universe, does not extend to non-Gaussian multivariate stable distributions. In this work, we provide the Laplace transform of bivariate stable distributions and its certain cut in the first quadrant. Given the lack of a closed-form likelihood function, we propose to estimate the parameters by means of Approximate Maximum Likelihood, a simulation-based method with desirable asymptotic properties. Simulation experiments and an application to truncated operational losses illustrate the applicability of the model.

economics,social sciences, mathematical methods,mathematics, interdisciplinary applications,business, finance
Modelling extremal dependence for operational risk by a bipartite graph

Oliver Kley,Claudia Klüppelberg,Sandra Paterlini

DOI: https://doi.org/10.1016/j.jbankfin.2020.105855

2020-08-01

Abstract:<p>We introduce a statistical model for operational losses based on heavy-tailed distributions and bipartite graphs, which captures the event type and business line structure of operational risk data. The model explicitly takes into account the Pareto tails of losses and the heterogeneous dependence structures between them. We then derive estimators and provide estimation methods for individual as well as aggregated tail risk, measured in terms of Value-at-Risk and Conditional-Tail-Expectation for very high confidence levels, and introduce also an asymptotically full capital allocation method for portfolio risk. Having access to real-world operational risk losses from the Italian banking system, we apply our model to these data, and carry out risk estimation in terms of the previously derived quantities. Simulation studies further reveal first that even with a small number of observations, the proposed estimation methods produce estimates that converge to the true asymptotic values, and second, that quantifying dependence by means of the empirical network has a big impact on estimates at both individual and aggregate level, as well as for capital allocations.</p>

economics,business, finance
An Analytical Portfolio Credit Risk Model Based on the Extended Binomial Distribution

Sven Fischer

DOI: https://doi.org/10.4236/jfrm.2019.83012

2019-01-01

Journal of Financial Risk Management

Abstract:The binomial distribution describes the probability of the number of successes for a fixed number of identical independent experiments, each with binary out-put. In real life, practical applications like portfolio credit risk management trials are not identical and have different realization probabilities. In addition to the number, the quantitative impacts of the respective outputs are also important. There exist no complete model-side implementations for the expansion of the binomial distribution, especially not in the case of specific quantitative parameters up to now. Here, a solution of this issue is described by the extended binomial distribution. The key for solving the problem lies in the use of bijection between the elementary events of the binomial distribution and the digit sequences of binary numbers. Based on the extended binomial distribution, an analytical portfolio credit risk model is described. The binomial distribution approach minimizes the approximation error in modeling. In particular, the edges of the loss distribution can be determined in a realistic manner. This analytical portfolio credit risk model is especially predestined for management of risk concentrations and tail risks.
An Extended Poisson Family of Life Distribution: A Unified Approach in Competitive and Complementary Risks

Pedro L. Ramos,Dipak K. Dey,Francisco Louzada,Victor H. Lachos

DOI: https://doi.org/10.48550/arXiv.1805.07672

2018-05-20

Abstract:In this paper, we introduce a new approach to generate flexible parametric families of distributions. These models arise on competitive and complementary risks scenario, in which the lifetime associated with a particular risk is not observable, rather, we observe only the minimum/maximum lifetime value among all risks. The latent variables have a zero truncated Poisson distribution. For the proposed family of distribution, the extra shape parameter has an important physical interpretation in the competing and complementary risks scenario. The mathematical properties and inferential procedures are discussed. The proposed approach is applied in some existing distributions in which it is fully illustrated by an important data set.

Applications
A Combination Model for Operational Risk Estimation in A Chinese Banking Industry Case

Jichuang Feng,Jianping Li,Lijun Gao,Zhongsheng Hua

DOI: https://doi.org/10.21314/jop.2012.106

2012-01-01

Abstract:The loss distribution approach is widely used in operational risk estimation. While the detailed form of frequency distribution does not significantly affect risk analysis, the choice of model in terms of severity often has a significant impact on operational risk measures. Usually, different heavy-tailed distributions, which have their own characteristics, are respectively used to quantify operational risk. In order to integrate characteristics of different heavy-tailed distributions and to reduce the uncertainty of the operational risk model, we propose a combination model to estimate operational risk in this paper. The model has three stages. First, we estimate operational risk by using different heavy-tailed distributions in the loss distribution approach framework. Second, the weights are decided according to certain criteria, and in this paper, p-values of the Kolmogorov-Smirnov goodness-of-fit test are used to decide the weight of each operational risk. Finally, the results obtained in the previous phases are combined in an integrated estimation to form the final result. We use a linear combination model to measure the operational risk capital of Chinese banks. We also compare the results measured using the combination model with those obtained from the basic indicator approach. Empirical analysis shows that this approach allows capital to be allocated in a more efficient way than the standard approach (only one heavy-tailed distribution used).
Composite Lognormal-T regression models with varying threshold and its insurance application

Girish Aradhye,Deepesh Bhati,George Tzougas

DOI: https://doi.org/10.48550/arXiv.2208.01262

2022-08-02

Abstract:Composite probability models have shown very promising results for modeling claim severity data comprised of small, moderate, and large losses. In this paper, we introduce three classes of parametric composite regression models with a varying threshold. We consider the Lognormal distribution for the head and the Burr, the Stoppa and the generalized log-Moyal (GlogM) distributions for the tail part of the composite family. Further, the Mode-Matching procedure has been utilized for the composition of the two densities. To capture the heterogeneous behavior of the policyholder's characteristics, covariates are introduced into the scale parameter of the tail distribution. Finally, the applicability of the proposed models has been shown using a real-world insurance data set.

Applications
Multinomial Backtesting of Distortion Risk Measures

Sören Bettels,Sojung Kim,Stefan Weber

2024-08-01

Abstract:We extend the scope of risk measures for which backtesting models are available by proposing a multinomial backtesting method for general distortion risk measures. The method relies on a stratification and randomization of risk levels. We illustrate the performance of our methods in numerical case studies.

Risk Management
Distributionally robust optimization approaches to credit risk management of corporate loan portfolios

Hansheng Sun,Roy H. Kwon

DOI: https://doi.org/10.21314/jcr.2024.008

IF: 0.645

2024-11-27

The Journal of Operational Risk

Abstract:A new approach to manage credit risk in financial institutions - the empirical divergence-based distributionally robust optimization - is proposed and shown to alleviate the challenges of sample sparsity and data uncertainty in credit risk modeling.

business, finance
Two maxentropic approaches to determine the probability density of compound risk losses

Erika Gomes-Gonçalves,Henryk Gzyl,Silvia Mayoral

DOI: https://doi.org/10.48550/arXiv.1411.5625

2014-11-22

Abstract:Here we present an application of two maxentropic procedures to determine the probability density distribution of compound sums of random variables, using only a finite number of empirically determined fractional moments. The two methods are the Standard method of Maximum Entropy (SME), and the method of Maximum Entropy in the Mean (MEM). We shall verify that the reconstructions obtained satisfy a variety of statistical quality criteria, and provide good estimations of VaR and TVaR, which are important measures for risk management purposes. We analyze the performance and robustness of these two procedures in several numerical examples, in which the frequency of losses is Poisson and the individual losses are lognormal random variables. As side product of the work, we obtain a rather accurate description of the density of the compound random variable. This is an extension of a previous application based on the Standard Maximum Entropy approach (SME) where the analytic form of the Laplace transform was available to a case in which only observed or simulated data is used. These approaches are also used to develop a procedure to determine the distribution of the individual losses through the knowledge of the total loss. Then, in the case of having only historical total losses, it is possible to decompound or disaggregate the random sums in its frequency/severity distributions, through a probabilistic inverse problem.

Risk Management,Methodology
Stability and Sample-based Approximations of Composite Stochastic Optimization Problems

Darinka Dentcheva,Yang Lin,Spiridon Penev

DOI: https://doi.org/10.48550/arXiv.2201.01338

2022-01-05

Abstract:Optimization under uncertainty and risk is indispensable in many practical situations. Our paper addresses stability of optimization problems using composite risk functionals which are subjected to measure perturbations. Our main focus is the asymptotic behavior of data-driven formulations with empirical or smoothing estimators such as kernels or wavelets applied to some or to all functions of the compositions. We analyze the properties of the new estimators and we establish strong law of large numbers, consistency, and bias reduction potential under fairly general assumptions. Our results are germane to risk-averse optimization and to data science in general.

Optimization and Control,Statistics Theory
On composite length-biased exponential-Pareto distribution: Properties, simulation, and application in actuarial science

Moulouk Halima Benchettah,Halim Zeghdoudi,Vinoth Raman

DOI: https://doi.org/10.3389/fams.2023.1137036

2023-03-20

Frontiers in Applied Mathematics and Statistics

Abstract:The composite length-biased exponential-Pareto (CLBEP) distribution is a new composite distribution that is introduced in this article. This model's probability density function, moments, and quantiles, among other statistical characteristics, are determined mathematically. The parameters' maximum-likelihood estimation and stochastic ordering are discussed. A comparison study with other new composite and conventional distributions is also included. Specifically, using two actual fire insurance data sets, the goodness of fit of this new model is contrasted with the composite exponential-Pareto, composite lognormal-Pareto, and composite Rayleigh-Pareto distributions (Algerian and Danish fire insurance losses). 2010 AMS subject classifications 62E10; 60E05.

English Else
Risk Aggregation and Allocation in the Presence of Systematic Risk via Stable Laws

Andrew Fleck,Edward Furman,Yang Shen

2024-10-19

Abstract:In order to properly manage risk, practitioners must understand the aggregate risks they are exposed to. Additionally, to properly price policies and calculate bonuses the relative riskiness of individual business units must be well understood. Certainly, Insurers and Financiers are interested in the properties of the sums of the risks they are exposed to and the dependence of risks therein. Realistic risk models however must account for a variety of phenomena: ill-defined moments, lack of elliptical dependence structures, excess kurtosis and highly heterogeneous marginals. Equally important is the concern over industry-wide systematic risks that can affect multiple business lines at once. Many techniques of varying sophistication have been developed with all or some of these problems in mind. We propose a modification to the classical individual risk model that allows us to model company-wide losses via the class of Multivariate Stable Distributions. Stable Distributions incorporate many of the unpleasant features required for a realistic risk model while maintaining tractable aggregation and dependence results. We additionally compute the Tail Conditional Expectation of aggregate risks within the model and the corresponding allocations.

Risk Management,Portfolio Management
Efficient exponential tilting with applications

Fuh, Cheng-Der

DOI: https://doi.org/10.1007/s11222-023-10374-5

IF: 2.3241

2024-01-14

Statistics and Computing

Abstract:To minimize the variance of Monte Carlo estimators, we develop a novel exponential embedding technique that extends the classical concept of sufficient statistics in importance sampling. Our method demonstrates bounded relative error and logarithmic efficiency when applied to normal and gamma distributions, especially in rare event scenarios. To illustrate this innovative technique, we address the problem of credit risk measurement in portfolios and present an efficient simulation algorithm to estimate the likelihood of significant portfolio losses, leveraging multi-factor models with a normal mixture copula. Finally, supported by comprehensive simulation studies, our approach offers a more effective and efficient way to simulate moderately rare events.

statistics & probability,computer science, theory & methods
Analytic Loss Distributional Approach Model for Operational Risk from the alpha-Stable Doubly Stochastic Compound Processes and Implications for Capital Allocation

Gareth W. Peters,Pavel Shevchenko,Mark Young,Wendy Yip

DOI: https://doi.org/10.48550/arXiv.1102.3582

2011-02-17

Abstract:Under the Basel II standards, the Operational Risk (OpRisk) advanced measurement approach is not prescriptive regarding the class of statistical model utilised to undertake capital estimation. It has however become well accepted to utlise a Loss Distributional Approach (LDA) paradigm to model the individual OpRisk loss process corresponding to the Basel II Business line/event type. In this paper we derive a novel class of doubly stochastic alpha-stable family LDA models. These models provide the ability to capture the heavy tailed loss process typical of OpRisk whilst also providing analytic expressions for the compound process annual loss density and distributions as well as the aggregated compound process annual loss models. In particular we develop models of the annual loss process in two scenarios. The first scenario considers the loss process with a stochastic intensity parameter, resulting in an inhomogeneous compound Poisson processes annually. The resulting arrival process of losses under such a model will have independent counts over increments within the year. The second scenario considers discretization of the annual loss process into monthly increments with dependent time increments as captured by a Binomial process with a stochastic probability of success changing annually. Each of these models will be coupled under an LDA framework with heavy-tailed severity models comprised of $\alpha$-stable severities for the loss amounts per loss event. In this paper we will derive analytic results for the annual loss distribution density and distribution under each of these models and study their properties.

Risk Management,Statistics Theory,Applications,Computation
Efficient algorithms for calculating risk measures and risk contributions in copula credit risk models

Zhenzhen Huang,Yue Kuen Kwok,Ziqing Xu

DOI: https://doi.org/10.1016/j.insmatheco.2024.01.005

2024-01-25

Insurance: Mathematics and Economics

Abstract:This paper innovates in the risk management of insurance and banking capital by exploring efficient, accurate, and reliable algorithms for evaluating risk measures and contributions in copula credit risk models. We propose a hybrid saddlepoint approximation algorithm, which leverages a synergy of nice analytical tractability from the saddlepoint approximation framework and efficient numerical integration from the Monte Carlo simulation. Notably, the numerical integration over the systematic risk factors is enhanced using three novel numerical techniques, namely, the mean shift technique, randomized quasi-Monte Carlo simulation, and scalar-proxied interpolation technique. We also enhance the exponential twisting and cross entropy algorithms via the use of interpolation and update rules of optimal parameters, respectively. Extensive numerical tests on computing risk measures and risk contributions were performed on various copula models with multiple risk factors. Our hybrid saddlepoint approximation method coupled with various enhanced numerical techniques is seen to exhibit a high level of efficiency, accuracy, and reliability when compared with existing importance sampling algorithms.

economics,social sciences, mathematical methods,mathematics, interdisciplinary applications,statistics & probability
Portfolio Optimisation within a Wasserstein Ball

Silvana Pesenti,Sebastian Jaimungal

DOI: https://doi.org/10.48550/arXiv.2012.04500

2022-06-20

Abstract:We study the problem of active portfolio management where an investor aims to outperform a benchmark strategy's risk profile while not deviating too far from it. Specifically, an investor considers alternative strategies whose terminal wealth lie within a Wasserstein ball surrounding a benchmark's -- being distributionally close -- and that have a specified dependence/copula -- tying state-by-state outcomes -- to it. The investor then chooses the alternative strategy that minimises a distortion risk measure of terminal wealth. In a general (complete) market model, we prove that an optimal dynamic strategy exists and provide its characterisation through the notion of isotonic projections. We further propose a simulation approach to calculate the optimal strategy's terminal wealth, making our approach applicable to a wide range of market models. Finally, we illustrate how investors with different copula and risk preferences invest and improve upon the benchmark using the Tail Value-at-Risk, inverse S-shaped, and lower- and upper-tail distortion risk measures as examples. We find that investors' optimal terminal wealth distribution has larger probability masses in regions that reduce their risk measure relative to the benchmark while preserving the benchmark's structure.

Mathematical Finance,Portfolio Management,Risk Management
Pricing multi-asset options with tempered stable distributions

Yunfei Xia,Michael Grabchak

DOI: https://doi.org/10.1186/s40854-024-00649-9

IF: 6.793

2024-08-23

Financial Innovation

Abstract:We derive methods for risk-neutral pricing of multi-asset options, when log-returns jointly follow a multivariate tempered stable distribution. These lead to processes that are more realistic than the better known Brownian motion and stable processes. Further, we introduce the diagonal tempered stable model, which is parsimonious but allows for rich dependence between assets. Here, the number of parameters only grows linearly as the dimension increases, which makes it tractable in higher dimensions and avoids the so-called "curse of dimensionality." As an illustration, we apply the model to price multi-asset options in two, three, and four dimensions. Detailed goodness-of-fit methods show that our model fits the data very well.

social sciences, mathematical methods,business, finance
Risk models from tree-structured Markov random fields following multivariate Poisson distributions

Hélène Cossette,Benjamin Côté,Alexandre Dubeau,Etienne Marceau

2024-12-01

Abstract:We propose risk models for a portfolio of risks, each following a compound Poisson distribution, with dependencies introduced through a family of tree-based Markov random fields with Poisson marginal distributions inspired in Côté et al. (2024b, <a class="link-https" data-arxiv-id="2408.13649" href="https://arxiv.org/abs/2408.13649">arXiv:2408.13649</a>). The diversity of tree topologies allows for the construction of risk models under several dependence schemes. We study the distribution of the random vector of risks and of the aggregate claim amount of the portfolio. We perform two risk management tasks: the assessment of the global risk of the portfolio and its allocation to each component. Numerical examples illustrate the findings and the efficiency of the computation methods developed throughout. We also show that the discussed family of Markov random fields is a subfamily of the multivariate Poisson distribution constructed through common shocks.

Methodology,Risk Management

Composite Tukey-type distributions with application to operational risk management

Multivariate Risk Models under Heavy-Tailed Risks

Data-driven and distribution-free estimation of tailed-related risks for GARCH models using composite asymmetric least squares regression

Some analytical results on bivariate stable distributions with an application in operational risk

Modelling extremal dependence for operational risk by a bipartite graph

An Analytical Portfolio Credit Risk Model Based on the Extended Binomial Distribution

An Extended Poisson Family of Life Distribution: A Unified Approach in Competitive and Complementary Risks

A Combination Model for Operational Risk Estimation in A Chinese Banking Industry Case

Composite Lognormal-T regression models with varying threshold and its insurance application

Multinomial Backtesting of Distortion Risk Measures

Distributionally robust optimization approaches to credit risk management of corporate loan portfolios

Two maxentropic approaches to determine the probability density of compound risk losses

Stability and Sample-based Approximations of Composite Stochastic Optimization Problems

On composite length-biased exponential-Pareto distribution: Properties, simulation, and application in actuarial science

Risk Aggregation and Allocation in the Presence of Systematic Risk via Stable Laws

Efficient exponential tilting with applications

Analytic Loss Distributional Approach Model for Operational Risk from the alpha-Stable Doubly Stochastic Compound Processes and Implications for Capital Allocation

Efficient algorithms for calculating risk measures and risk contributions in copula credit risk models

Portfolio Optimisation within a Wasserstein Ball

Pricing multi-asset options with tempered stable distributions

Risk models from tree-structured Markov random fields following multivariate Poisson distributions