The Determinants of Credit Default Swap Premia and the Use of Machine Learning Techniques for their Estimation

Feser, Julian Alexander and Broby, Daniel (2020) The Determinants of Credit Default Swap Premia and the Use of Machine Learning Techniques for their Estimation. Discussion paper. University of Strathclyde, Glasgow.

[thumbnail of Feser-Broby-2020-The-determinants-of-credit-default-swap-premia-and-the-use-of-machine-learning]
Preview
Text. Filename: Feser_Broby_2020_The_determinants_of_credit_default_swap_premia_and_the_use_of_machine_learning.pdf
Final Published Version
License: Creative Commons Attribution 3.0 logo

Download (300kB)| Preview

Abstract

This paper examines the determinants of Credit Default Swap premia. It also explores the use of Machine Learning techniques for their estimation. We address default risk, counter-party risk and liquidity risk. We discuss these in the context of yield curves, maturity and volatility. The insights gained are used to illustrate how the use of technology can provide more timely, efficient and informative valuations. We recommend the use of support vector and artificial neural networks (supervised learning models), as well as principal component analysis. Combined with standardized electronic processing and central clearing of trade, we suggest that this will enhance the depth of CDS markets. At the same time, Machine Learning can also aid the understanding of the various premia. We conclude that the application of Artificial Intelligence can add significant economic value to banking operations.

ORCID iDs

Feser, Julian Alexander and Broby, Daniel ORCID logoORCID: https://orcid.org/0000-0001-5482-0766;