Ursulenko G. Banking risk modeling under condition of Basel II.

The thesis is devoted to the study of modern approaches to the banking risk definition and modeling. The economic nature of banking risks are analyzed in terms of domestic and international law, proved the importance of modeling in the banking sector. The thesis deals with the models based on the Bayesian approach to evaluating and determining the probability of default and minimum capital requirements. In the framework of the Bayesian approach, we consider two ways of modeling credit risk: an approach based on Monte Carlo Markov Chains approximation (MCMC) and Integrated Netscape Laplace Approximation (INLA). On the basis of the INLA approach the dissertation depicts the technique of Bayesian analysis for credit risk modeling (the use of structural models with noise). The thesis analyses using VaR for market risk modeling and evaluating. Depending on the probability distribution of risk factors and type of functional relationship between the change in value of the portfolio and changes in risk factors considered basic methods and models for calculating VaR: covariation model; GARCH-model, extreme theory model.