Master of Science (MS)
This work, dealing with the correlation between subportfolios in more complex portfolios, begins with a brief survey of the necessary theoretical background. The basic statistical and probabilistic concepts are reviewed. The notion of copulas is introduced along with the fundamental theorem of Sklar. After this background a numerical procedure and code are developed for correlated defaults in multiple correlated portfolio. Further on, interesting results regarding the impact of changes in correlation on the portfolio performance are investigated in the simulations. The most valuable observations regarding the expected default ratios of two subportfolios considered jointly are presented and explained with particular care. These observations are compared with theoretical results. The sensitivity of the tranche losses to correlation parameters is examined carefully. The work is concluded with a brief summary of the most significant observations and their possible impact on portfolio performance.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Lodygowski, Adam, "Correlation of defaults in complex portfolios using copula techniques" (2010). LSU Master's Theses. 617.