'Too central to fail' firms in bi-layered financial networks: linkages in the US corporate bond and stock markets. Issue 5 (4th May 2022)

• Record Type: Journal Article
• Authors: Mishra, Abinash; Srivastava, Pranjal; Chakrabarti, Anindya S.
• Is Part Of: Quantitative finance. Volume 22:Issue 5(2022)
• Page Start: 943
• Page End: 971
• Publication Date: 2022-05-04
• Subjects: Network centrality -- Risk management -- Vulnerability -- Stocks and bonds -- Too central to fail; G01 -- G32 -- G17 -- G18; Finance -- Periodicals; Business mathematics -- Periodicals; Finance -- Mathematical models -- Periodicals; Investments -- Mathematics -- Periodicals; Economics -- Periodicals; Finances -- Modèles mathématiques -- Périodiques; 332.015118
• DOI: 10.1080/14697688.2021.2006281
• Languages: English
• ISSNs: 1469-7688

Abstract

Abstract : Several measures have been recently developed in the financial networks literature to quantify the vulnerabilities of firms in specific markets for risk management. However, firms are often active in multiple asset markets simultaneously with potentially different degrees of vulnerabilities across markets. One can hypothesize that since assets are backed by similar firm-level fundamentals, the vulnerability measures across markets would be highly correlated. In this paper, we present two results based on the empirical correlation of network-based measures of vulnerabilities by studying US firms' asset returns that are active in stock as well as corporate bond markets. First, the magnitude of the relationship, while positive and significant, is not very large. Quantitatively, a unit percent increase in vulnerability measure of a firm in the stock market is associated with 0.15% increase in the vulnerability in the bond market. Second, the vulnerability of the firms in the stock market is negatively related to firm size proxied by market capitalization, indicating that 'too-big-to-fail' firms tend to be 'too-central-to fail'. By adopting instrumental variables, we show that the coefficient has a higher magnitude. The results are robust with respect to choices of asset classes, maturity horizons, model selection, time length of the data, as well as controlling for return sensitivities to market-level factors.