Slow- and fast-moving information content of CDS spreads: new endogenous systematic factors

Abstract

This paper proposes two new Credit Default Swap (CDS) endogenous systematic factors constructed from peer-CDS information. The factors capture slow-moving credit risk information, as well as fast-moving newly arrived market information embedded in the most recent CDS quotes. Using a sample of U.S. non-financial listed firms from 2002 to 2011, we find that these two endogenous systematic factors dominate firm-specific factors and other widely known systematic factors in insample and out-of-sample CDS spread predictions.

Corporate Credit Default Swap Systematic Factors

“…Additional 5 firm-specific variables -cash holding (CASHMTA), CashRatio, CDS highminus-low (CDSHL), CDSSlope, and stock Amihud (StoAmihud) -are shown to explain the contemporaneous changes of the CDS srpeads, most of which are related to illiquidity. This highlights the illiquidity impact on individual CDS spread changes (Coró et al, 2013;Lin et al, 2019). Albeit, overall merely one-third of the firm-specific variables included in this study show statistical significance in explaining CDS spread movement, which, as we argue, casts doubt on the importance of the firm-specific variables.…”

“…The Amihud (2002) measure of the underlying stock over one year. , Lin et al (2019). CDSAmihud The Amihud (2002) measure of the CDS spreads over one year.…”

“…2 Houweling and Vorst (2005) compared the model prices and the market prices of the CDS spread and conclude that the theoretical pricing models were in general working fine. But Lin et al (2019) document that the CreditGrade model, a CDS pricing model based on Merton (1974), only captures 9% of the monthly variation of the CDS spreads.…”

“…overall, rating, and sector levels) all contribute to the CDS price variations, while the prior studies only cover parts of the peer information, e.g. Galil et al (2014) (sector, overall) and Lin et al (2019)…”

Significance, relevance and explainability in the machine learning age: an econometrics and financial data science perspective

“…Specifically, the papers in this special issue of The European Journal of Finance inform academics, practitioners and policymakers regarding developments in Financial Data Science and Econometrics and provide insight on how cutting-edge data and/or cutting-edge statistical techniques can be applied to enhance our understanding of financial phenomena. Three broad themes are covered in these research papers: financial forecasting and the related statistical significance (McMillan 2020; Meligkotsidou et al 2020), financial risk modelling and related statistical relevance (Kim, Park, and Yoon 2020;Lin, Kolokolova, and…”