|
|
Determining the Exposition Measure of the Credit and Market Risk Using the VaR Methods
Friedrichová, Andrea ; Stařík, David (advisor) ; Benková, Markéta (referee)
The thesis examines the share of market and credit exposition on the total rate of risk of an equity index. The paper describes models for estimation of market risk using the Value-at-Risk methods, which are the parametric approach, the historical simulation and the Monte Carlo simulation. Further, it describes the estimation of credit risk using the Value-at-Risk. The main goal is to descibe and then to adopt in practise three methods for calculating integrated VaR: integrated VaR model based on historical data, integrated VaR with regard to the covariance between market and credit risk and integrated VaR based on the parametric approach to VaR. These methods are applied to selected equities of the index S&P 500 and compared.
|
|
| |
|
|
Actuarial approach to credit risk modelling
Benešová, Milena ; Mandl, Petr (referee) ; Benková, Markéta (advisor)
This thesis deals with one of the models for the credit risk measurement - the model CreditRisk+. The theoretical part describes the theory which is the basis for this model. Further, the thesis demonstrates an applicative example of calculation distribution of default losses. The model uses Poisson distribution as the distribution of the number of defaults from this we can proceed to the distribution of default losses which is output from this model. The theoretical part also presents two variants of this model. The first of this variant is the calculation of the distribution of default losses with fixed default rates. The main asset of this model is the second variant which calculates with the variable default rates. The applied part deals with the recurrence relation which is described with the model-makers. This thesis deals with the combination of CreditRisk+ with the another model known as CreditMetrics, too. The calculation is realized on the basis of Monte Carlo's simulation of the future portfolio. The aim of this part is to demonstrate how this model is applicable in practise.
|
|
|
Calculation of the Credit Value at Risk
Zamazal, David ; Mandl, Petr (referee) ; Benková, Markéta (advisor)
Thesis describes calculation of the credit value at risk for portfolio composed of traditional bank loans. The risk is measured by incurred expected and unexpected losses at the end of some time horizon. Thesis is splitted into two parts - theoretical part and computational part. The most known and most widely used models are described in the first part, in conjunction with definition of their main input parameters - probability of default, exposure at default, loss given default and correlation between debtors. Detailed theoretical description of two chosen methods comes after - CreditMetrics method and Vasicek's method. The examined portfolio is characterized in the computational part, along with other input parameters, essential for evaluation. Then model implementation into software Mathematica is described, evaluation run and the results. Eventually both methods are compared.
|
|
|
Credit Risk Valuation
Pleška, Martin ; Benková, Markéta (referee) ; Charamza, Pavel (advisor)
According to the rules stated in the Basel II document banks are obliged to calculate risk capital on the basis of expected value of credit risk and in particular on the basis of some of its characteristics among which is Value at Risk (VaR) also ranked. It can be calculated for example by the method stated in Creditmetrics paper. In this thesis we will focus on this method of calculation of VaR which is considered to be a measure of credit risk. Determination of expected value of portfolio which credit risk we are concerned about is in this paper demonstrated by two methods. First one is the method of discounted cash řow and the second one is the method of risk costs. Estimations of VaR are being performed through the use of simulation of distribution of the value of the portfolio. The work is amended by a particular calculation with real data.
|
guest ::
RSS feed.