Sigorta Sektörünün Kredi Portföy Risk Modeli İle Değerlendirilmesi
xmlui.mirage2.itemSummaryView.MetaDataShow full item record
The credit risk, which emerges as a concept of default risk, can be defined as the risk that any liability of the borrower can not be fulfilled. Default cases arise out of the situations of delayed payments, the presence of non-payment periods or bankruptcy. The existence of default cases pose a considerable risk to the lender. For this reason, the lender must carry out risk management in order to protect himself. In the event of possible default, the lender should compute the loss and determine the economic capital accordingly. Economic capital gives the amount between expected and unexpected loss. While it is easy to calculate the expected loss, it is difficult to calculate the unexpected loss. Therefore, credit risk models have been developed and a resource has been created for financial institutions. Sometimes financial institutions use these available methods and sometimes produce their own internal models. While credit risk was measured only on credit basis in previous periods; it is now mostly measured on portfolio basis in the new period. In the portfolio-based measures, taking into account the relationship of borrowers with each other, a general portfolio loss distribution is found, so that amount of economic capital, that should be allocated for risk purposes, can be computed more accurately. In portfolio-based credit risk models, default probabilities, transition matrices, correlations and recovery rates are used depending on model selection. In addition, the portfolio manager can determine, depending on the risk contributions of the borrowers, the continuity of borrowers in the portfolio. When it is needed, the portfolio manager can diversify the portfolio and also play with the amount of concentration. Portfolio-based credit risk models used in the literature are CreditRisk+ model with an actuarial approach, CreditPortfolioView model with a macroeconomic approach and CreditMetrics and KMV models based on asset value. These models show similarities and differences in some respects. In this study, information about models is given, and potential similarities and differences are also mentioned. In addition, we conduct a numerical study in Turkey related to the companies which are investigating non-life insurance, with the data provided by the Insurance Association of Turkey. In the study, a portfolio consisting of companies was considered and a portfolio risk analysis was made. Bankruptcy probabilities of companies are estimated and simulations are created based on these probabilities. For the analysis, the simulation technique with the logic of the CreditMetrics method is used and the CreditRisk+ method is applied together with this technique. Thus, the loss distribution of the insurance sector is obtained. In addition, both the risk contributions of companies to the insurance sector and the amount of capital required to hold them in their hands have been determined. Thus, a general observation has been made about the insurance sector and the state of the sector has been learned in accordance with assumptions.