Credit risk management is mainly concerned with the calculation and controlling of the potential losses that banks may face if their obligors default or the obligors' creditworthiness deteriorates. As the calculation of the capital reserve requirements is based on the unexpected losses of one period, most regulators do not focus on the evolution of the credit portfolio over multiple periods. On the opposite side, banks often need to take into account the expected losses over multiple years when they are considering long-term investment decisions or assessing potential losses from an accounting perspective. As there is no closed-form solution of the loss distribution of multi-period credit risk models, it needs to be quantified using scenario analysis or Monte Carlo simulations. This Master's thesis introduces a generalized multi-period credit risk model and uses Taylor series expansion to estimate its loss distribution. Moreover, we show how to estimate the model parameters with maximum likelihood estimation and an integrated Kalman Filter. We demontrate our model based on the Normal Inverse Gaussian distribution and compare its results to the multi-period Vasicek model. Our numerical results show that the multi-period loss approximation is more accurate during times of economic recession and for credit portfolios with a high probability of default or with a strongly autocorrelated common risk factor. With regards to computational effort and numerical accuracy, we conclude that the second-order approximation is preferable. Furthermore, the implementation of the model parameter estimation proved challenging for the Normal Inverse Gaussian distribution. It was significantly less accurate compared to the multi-period Vasicek model which may be explained by the complexity of the log-likelihood function and non-identifiability of the model.     «

Credit risk management is mainly concerned with the calculation and controlling of the potential losses that banks may face if their obligors default or the obligors' creditworthiness deteriorates. As the calculation of the capital reserve requirements is based on the unexpected losses of one period, most regulators do not focus on the evolution of the credit portfolio over multiple periods. On the opposite side, banks often need to take into account the expected losses over multiple years when...     »