Abstract:We aim to analyze the behaviour of a finite-time stochastic system, whose model is not available, in the context of more rare and harmful outcomes. Standard estimators are not effective in making predictions about such outcomes due to their rarity. Instead, we use Extreme Value Theory (EVT), the theory of the long-term behaviour of normalized maxima of random variables. We quantify risk using the upper-semideviation $\rho(Y) = E(\max\{Y - \mu,0\})$ of an integrable random variable $Y$ with mean $\mu = E(Y)$. $\rho(Y)$ is the risk-aware part of the common mean-upper-semideviation functional $\mu + \lambda \rho(Y)$ with $\lambda \in [0,1]$. To assess more rare and harmful outcomes, we propose an EVT-based estimator for $\rho(Y)$ in a given fraction of the worst cases. We show that our estimator enjoys a closed-form representation in terms of the popular conditional value-at-risk functional. In experiments, we illustrate the extrapolation power of our estimator using a small number of i.i.d. samples ($<$50). Our approach is useful for estimating the risk of finite-time systems when models are inaccessible and data collection is expensive. The numerical complexity does not grow with the size of the state space.

Some applications and methods of large deviations in finance and insurance

Multivariate Risk Models under Heavy-Tailed Risks

Precise Large Deviations for the Actual Aggregate Loss Process

Large deviations for a class of multivariate heavy-tailed risk processes used in insurance and finance

Large deviations for random sums of differences between two sequences of random variables with applications to risk theory

Large Deviations for Sums of Claims in a General Renewal Risk Model with the Regression Dependent Structure

Extreme Value Theorems of Uncertain Process with Application to Insurance Risk Model

One Method of Application of Extreme Value Theory to Computing Value-at-Risk

Precise large deviations of the net loss process in a non-standard two-dimensional risk model

Large Deviations for Exchangeable Observations with Applications

Large Deviation of Perturbed Risk Model with Heavy-Tailed Claims and Variable Premium Rate

Asymptotic Multivariate Finite-time Ruin Probability with Statistically Dependent Heavy-tailed Claims

Precise large deviations of aggregate claims in a risk model with regression-type size-dependence

Large Deviations for Inhomogeneous Systems

Precise large deviations of some risk objectives related to the net loss process in two nonstandard risk models

Two examples of non strictly convex large deviations

Large Deviations for Heavy-Tailed Random Sums in Compound Renewal Model

Large Portfolio Asymptotics for Loss From Default

Toward Scalable Risk Analysis for Stochastic Systems Using Extreme Value Theory

The Entropy Method in Large Deviation Theory

AN EXTREME-VALUE THEORY APPROXIMATION SCHEME IN REINSURANCE AND INSURANCE-LINKED SECURITIES