Stanovení kreditního rizika portfolia obligací obchodovaných na LSE dle metody CreditMetrics

Abstract

Thesis is focused on credit risk determination of bond portfolio traded on LSE through CreditMetrics methodology. The whole thesis is separated into two main parts. In the first (theoretical) part is characterized credit risk together with tools and models used for its quantification. In the second (practical) part is applied theoretical knowledge about CreditMetrics on the two portfolios consisting of nine company obligations and one government obligation. Basis of credit risk calculation is determination of the correlated yields through Monte Carlo simulation. To these yields are assigned ratings according to the inferred transition thresholds. Knowledge of ratings enables determination of the portfolio values for 10 000 scenarios, which are used for distribution function compilation and calculation of marginal standard deviation representing the risk degree. Results are presented in written, tabular and also graphical form and at the end they are concluded in separate chapter. Government obligation is considered to be the least risky one. The riskiness in the case of company obligations is influenced mainly by their credit exposure, which is affected by bond face value.