loading page

Copula functions in Credit Metrics' VaR estimation
  • Capstone Project by Shynggys Magzanov,
  • Supervisor: Dr.Dongming Wei, Second Reader: Dr.Zhenisbek Assylbekov
Capstone Project by Shynggys Magzanov
Nazarbayev University
Author Profile
Supervisor: Dr.Dongming Wei, Second Reader: Dr.Zhenisbek Assylbekov

Abstract

Credit risk modelling of a portfolio of exposures is essential part of activity of every financial institution. However this procedure is complicated since the joint behavior of chosen exposures must be known. In this paper Value at Risk percentile of the portfolio consisting of three corporate bonds issued by Lukoil, Gazprom and Norilsk Nickel was estimated at three different significance levels within the frame of Credit Metrics approach proposed by J.P.Morgan. Following the Asset value model, Monte-Carlo simulations were performed to obtain possible portfolio values in one year time horizon. Where the joint distribution of asset returns of three companies was constructed by means of pair-copula construction method discussed in Aas, Czado, Frigessi,Bakken (2009). Results reveal that for particular portfolio of bonds at 90%, 95% and 99% confidence levels the value of our portfolio will not fall below 2057.915 ,1798.117 and 1375.011 dollars respectively.