Facoltà di scienze economiche

Essays on credit risk

Borghi, Matteo ; Barone-Adesi, Giovanni (Dir.)

Thèse de doctorat : Università della Svizzera italiana, 2013 ; 2013ECO005.

The work investigates two major topics: the presence of a systematic and an idiosyncratic component in CDS spreads and the credit spread puzzle. We verify that a systematic factor is priced in the cross-section of CDS returns. We also notice that the systematic component of risk increases after the financial crisis. We finally verify that the fraction of systematic risk is not the same in... More

Add to personal list
    Summary
    The work investigates two major topics: the presence of a systematic and an idiosyncratic component in CDS spreads and the credit spread puzzle. We verify that a systematic factor is priced in the cross-section of CDS returns. We also notice that the systematic component of risk increases after the financial crisis. We finally verify that the fraction of systematic risk is not the same in different industrial sectors. In particular, more cyclical and systemic sectors show a much larger impact of the systematic factor. Regarding the second topic, we extend the literature proposing a bivariate state space model and verify that it actually improves the performances of standard inversion techniques in explaining the observed credit spreads. The improvement is particularly significant during the crisis period, characterized by a larger noise contaminating the observed equity price and equity volatility. This supports the ability of the state space model to remove the noise component and to produce better estimates of the asset value of the company and, consequently, more accurate predictions of spreads. In the last chapter we identify some explicit drivers for the noise postulated in the second paper. In particular, we verify that the errors produced by structural credit risk models significantly depend on liquidity indicators and that their explained variability is not negligible. We finally verify that the errors left by both structural variables and liquidity indicators are strongly correlated with market-wide measures of limits of arbitrage and/or deleveraging pressures.