Credit risk monitoring in the Czech banking sector : Early warning model

The aim of my thesis is in the first place to show how to deal with a credit risk, and which tools the Czech banking sector uses to minimize it (based on the adequate literature and own experience). In the second place, the aim is to find out the reliable logit model estimating the probability of default during the short period based on available data (in the time of economic crisis in the Czech environment). In the first part of this thesis I am describing the development of Non-performing loans before and during the current financial crisis together with the results of the CNB's stress tests. Next chapter describes the credit risk with the emphasis on the credit monitoring, including the most frequently used monitoring tools. Practical part turns us to the most important EWM model. The strictly confidential banking data (credit account turnovers, credit contract), together with data from the financial statements and CRU registry are the inputs to the Model. The Model should work as an early warning signal detection thanks to the estimate of probability of default (more specifically the watch loan classification or worse) during the next three months