The aim of this research is to produce a model allowing me to estimate the credit risks in the aggregate and the sectors levels of the Swedish economy in response to the evaluation of key macroeconomic variables. In order to estimate the credit risk models for the Swedish economy, one-factor models were used and the employed data were covering the period from 2003 to 2011. One factor models’ estimations for the sectors facilitate a comparison of default rates’ determiners between different sectors. The analyze part of the thesis starts with the estimation of the credit risk model at the aggregate economy level and it follows by the estimation of the models for different sectors. Ten different sectors are analyzed and for all sectors, the default rate models are produced. Furthermore, the paper presents some examples of applying the estimated models to macro stress testing. The findings demonstrate that in the transport and in the sector others, the most significant macroeconomic indicators were GDP, interest rates and repo rates. But, in all other sectors: GDP, interest rates and inflation rates showed the highest significant results. All coefficients were significant at the 5 % confidence level either in aggregate level or in sectors level. The interest rates showed positive relations with the default rates while the GDP and the inflation rates showed opposite relations. Reciprocal analyzes of the sectors indicated that compared to other sectors, the default rates in the financial sector strongly depended on the GDP and in the construction sector it weakly depended on inflation rates. In addition, the credit risks were varying between the sectors. At the education and the sector others, default rates were low, fluctuated between 0 and 0.05%. In contrast, at the manufacturing, the wholesale, the transportation, and the finance sectors the default rates were very high. It fluctuated between 0.03% and 0.16%. Finally, estimated models were used for the sensitive analyze of default rates by creating shocks over the independent variables. So, these calculations provided that, the default rates in financial activities sector were the most sensible sector during the shock at the GDP and the default rates in the construction sector were the most insensible ones during the shock at the interest rates and the inflation rates. To conclude, the results of this thesis can help understand the relationship between credit risk and macroeconomic indicators. This research provides important findings on how the macroeconomic indicators influence the default rates of Swedish economy either at the aggregate or at the sectors level. The calculated models can be used for the default rates’ prediction or stress testing.

The thesis evaluates relationship between probability of default of non-financial corporations and households and evolution of macroeconomic environment. This work contributes to the literature of credit risk proving importance of macroeconomic variables in determining the PDs both on aggregate level and for sector of non-financial corporations and sector of households in the Czech Republic. Evaluation of an impact of the recent financial crisis on the PDs are done by employing latent factor model and FAVAR model on monthly data of non-performing loans and other macroeconomic variables covering the period 01/2002-06/2013. Finally, an ability to forecast and fit the data of FAVAR model and one factor latent model are compared. The comparison indicates that latent factor model should be more appropriate than FAVAR model.

In this thesis we use nonstandard methods for the valuation of derivatives on electricity. We model the dynamics of electricity spot price as mean reverting processes on the hyperfinite binomial tree and by switching to the risk-neutral world we derive analytical formulas for the price of forward contracts. Both of our models are fitted to the German electricity market and forward price predictions are compared with forward products traded on the exchange. We conclude that both the Ornstein-Uhlenbeck and the Schwartz one factor model fit long-term forward contracts well while our prediction results for short-term forward prod- ucts are not conclusive due to low liquidity and alternative approaches might be suitable. 1