Analyzing credit spread changes using explainable artificial intelligence

We compare linear regression, local polynomial regression and selected machine learning methods for modeling credit spread changes. Using partial dependence plots (PDPs) and H-statistic, we find that the outperformance of machine learning models compared to regression ones is mostly attributable to...

Full description

Saved in:
Bibliographic Details
Published inInternational review of financial analysis Vol. 94; p. 103315
Main Authors Heger, Julia, Min, Aleksey, Zagst, Rudi
Format Journal Article
LanguageEnglish
Published Elsevier Inc 01.07.2024
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:We compare linear regression, local polynomial regression and selected machine learning methods for modeling credit spread changes. Using partial dependence plots (PDPs) and H-statistic, we find that the outperformance of machine learning models compared to regression ones is mostly attributable to complex non-linearities and not to interactions. The PDPs are additionally used to perform a factor hedging. For the first time, credit spread changes are decomposed by applying SHapley Additive exPlanation (SHAP) values. The proposed framework is applied to US and Euro Area corporate and covered bond credit spread changes of different maturities to quantify the influence of several macroeconomic and financial variables. Despite several commonalities between the decompositions of US and Euro Area credit spread changes, we also observe some differences — particularly related to the impact of certain explanatory variables during crisis periods. •We analyze the determinants of credit spread changes using machine learning methods.•We apply explainable artificial intelligence techniques to examine if the outstanding performance of the machine learning models is attributable to non-linear relationships or interactions.•We make use of partial dependence plots for hedging purposes.•We propose a novel approach based on SHAP values for quantifying the influence of different macroeconomic and financial variables on credit spread changes.•We validate the proposed methodologies on US and Euro Area corporate and covered bond spread changes of different maturities.
ISSN:1057-5219
1873-8079
DOI:10.1016/j.irfa.2024.103315