(Research Project) M.B.A.: Master of Financial Risk Management
In this paper, we assess the effects of estimation error due to the impact of noisy input parameters in portfolio credit risk modelling by using Monte-Carlo simulations. We employ the methodology used in L?ffler (2003) but apply different dataset to form two new portfolios: obligors with investment-grade credit rating and obligors with speculative-grade credit rating. The four sources of estimation risk are considered for each portfolio: default rate uncertainty only, recovery rate uncertainty only, correlation uncertainty only, and the three sources of uncertainty together. The resulting estimation error in the distribution of portfolio losses is considerable. The paper also shows that different credit datasets could result in different biases in value at risk (VaR) estimations in each portfolio.
Research Project (M.B.A.) - Simon Fraser University
Copyright is held by the author.
Supervisor or Senior Supervisor
Thesis advisor: Klein, Peter
Member of collection