Resumo

DAVILA ARAGON, Griselda; ORTIZ ARANGO, Francisco  e  CRUZ ARANDA, Fernando. Operational value at risk by bayesian networks for a financial firm. Contad. Adm [online]. 2016, vol.61, n.1, pp.176-201. ISSN 0186-1042.  https://doi.org/10.1016/j.cya.2015.09.009.

The aim of this paper is to outline the methodology based on the use of Bayesian networks (BN) to identify and quantify operational risk (OR) factors associated with processing financial transactions through electronic means in a financial company. BN model developed is exemplified with data from simulated events equivalent to six years period, from information provided by experts in this type of process. This represents one of the main advantages of using BR, they allow modeling the cause-effect relationships between different OR factors. Finally operational value at risk (OpVaR) for the example is calculated, where interacting factors that are not considered in the traditional model are incorporated, providing better conditions of credibility to this value.

Palavras-chave : Operational risk; Bayesian networks; Electronic transactions; Derivation trees.