تخمين زيان به ‌شرط نكول با در نظر گرفتن ساختار بدهي و نقدشوندگي وثيقه با استفاده از چهارچوب مدل‌هاي ساختاري و شبيه‌سازي مونت ‌كارلو

Estimating Loss Given Default Considering Firm’s Debt Structure an‎d Collateral Liquidity using structural models framework an‎d Monte Carlo simulation approach

مديريت

Abstract Due to the importance of credit risk management for financial institutions, especially banks, this thesis focuses on estimating the Expected Loss Given Default (ELGD). In this study, a structural model for estimating the ELGD is introduced, utilizing a Monte Carlo simulation approach to provide a method for estimating this variable for debt, considering diverse debt structures an‎d collateral liquidity of secure debts. The expected LGD is estimated for each debt in four debt structures assuming two types of collateral (liquid an‎d less liquid). For this purpose, it is assumed that the market price of the firm’s assets follows a Jump-Diffusion process, market price of liquid collateral follows a Geometric Brownian Motion process, an‎d market price of less liquid collateral follows a Trending Ornstein-Uhlenbeck process. The penalty for illiquidity is also considered for less liquid collateral. Collaterals are chosen from the firm’s tangible assets. The results indicate that the liquidity of collateral an‎d debt structure significantly affect the estimated ELGD. Specifically, using liquid collateral as debt guarantee, reduces the ELGD of both the secured debt an‎d unsecured debt. Statistical tests show that both the ETLGD of secured debt an‎d the ELGD of unsecured debt are influenced by the debt structure. Additionally, the firm level ELGD values indicate that this variable is also affected by collateral liquidity an‎d debt structures. A firm that uses only less liquid collateral to secure its debts has a higher expected firm level LGD compared to other debt structures. Sensitivity analysis of the parameters of the firmʹs total asset stochastic process showed that the expected loss given default (ELGD) of unsecured debt an‎d the expected total LGD of secured debt are decreasing functions of the drift parameter an‎d increasing functions of the diffusion an‎d jump intensity parameters. However, the expected LGD of secured debt, which only has access to collateral at the time of default, increases with an increase in the drift parameter in the company’s asset dynamics. Among the other studied influencing factors, the loan-to-value ratio plays a very important role in the expected LGD, especially for secured debts. As an empirical application, the proposed method was used to estimate the ELGD for eleven firms listed on the Tehran Stock Exchange which had sufficient information for estimating LGD, both at debt an‎d firm level. Keywords: Expected Loss Given Default, Debt structure, Collateral liquidity, Loan-To-Value ratio

5 فصل

160514