OPERATIONAL IT RISK FORECASTING BASED ON EXTREME VALUE THEORY
https://doi.org/ 10.18384/2310-6646-2018-2-145-154
Abstract
This article considers the mathematical economic model for IT operational risk estimation, which is based on extreme value theory. This model allows of predicting maximum possible loss caused by IT incidents in the context of multiple releases of automated banking systems. The model is based on the assumption that catastrophic losses from IT incidents are distributed as Fisher-Tippet distribution. The paper provides two different techniques for parameter estimation for the Fisher-Tippet distribution when statistical data on operational risk events is known. Calculations are made by means of R programming language for confirmation of the results of the work. The model is validated by means of the Kupiec test. The author describes advantages and disdvantages of the use of extreme value theory to estimate operational risks.
References
1. Зарядов И.С. Введение в статистический пакет R: типы переменных, структуры данных, чтение и запись информации, графика. М.: Издательство Российского университета дружбы народов, 2010. 207 с.
2. Петросян Г.С. Методы анализа операционных рисков при управлении релизами банковских информационных систем // Фундаментальные исследования. 2017. № 11-1. С. 108-113.
3. Шведов А.С. Теория вероятностей и математическая статистика: промежуточный уровень. М.: ИД Высшей школы экономики, 2016. 280 с.
4. Banking Banana Skins 2015. The CSFI survey of bank risk [Электронный ресурс] // PWC: [сайт]. URL: https://www.pwc.com/gx/en/financial-services/pdf/Banking-banana-skins-2015-final.pdf (дата обращения: 13.01.2018).
5. Novak S.Y. Extreme Value Methods with Applications to Finance. Florida: CRC Press, 2011. 399 p.
6. Operational risk loss data for banks submitted in 2016 [Электронный ресурс] // Managingrisktogether: [сайт]. URL: https://managingrisktogether.orx.org/research/beyond-headlines (дата обращения: 13.01.2018).
7. Scandizzo S. The Validation of Risk Models: A Handbook for Practitioners. New York: Palgrave Macmillan, 2016. 242 p.
8. Wickham H. R for Data Science: Import, Tidy, Transform, Visualize, and Model Data. Canada: O’Reilly Media, 2016. 522 p.
9. Yan J., Dey D.K. Extreme Value Modeling and Risk Analysis: Methods and Applications. Florida: CRC Press, 2016. 540 p.
Review
Views:
89
Email this article 