Empirical Studies of Credit Spreads and Ratings
Author(s)
Wu, Shi
Type
Thesis
Abstract
The credit markets experienced fundamental changes during the last
two decades. Corporate debt volumes have expanded rapidly making such
debt one of the largest asset classes in financial markets. Structured credit
products revolutionized banking practices and transformed the framework
of credit risk management. Credit risk modeling, valuation and rating have,
in consequence, been extensively studied. Many issues remain controversial
for both academics and industry practitioners, however, especially in the
wake of the recent financial crisis.
This thesis contributes to the existing literature in four ways. First,
a new approach is introduced for jointly estimating rating-specific bond
spread term structures. The approach achieves fitting accuracy comparable
to the more standard extended Nelson-Siegel technique meanwhile yielding
well-behaved and stable term structures which are appropriately ordered for
different rating classes.
Second, a ‘clinical’ study is presented of the evolution of Asset Backed
Security (ABS) valuations conditional on ratings during two crisis periods,
namely the 2007-2009 crisis and the earlier collapse of ABS secured against
Manufactured Housing loans of 2002-2003. The study focusses on the market’s
reactions to different pricing factors in crisis time including liquidity
and risk premiums.
Third, we examine predictable fluctuations in agency ratings behavior
over the business cycle and propose simple, tractable ways of parameterizing
time-varying rating transition matrices.
Fourth, we examine the dependence of corporate bond market risk premiums
on industrial sector. The analysis sheds light on the hypothesis that
the cross-sectional pattern of risk premiums reflects the degree to which
bond issuers operate in cyclically sensitive industries. By adjusting sectorspecific
spreads for time-varying and sector specific expected losses and for
liquidity, we show that spreads and, even more so, risk premiums are related
to the cyclicality of the obligor’s industry.
two decades. Corporate debt volumes have expanded rapidly making such
debt one of the largest asset classes in financial markets. Structured credit
products revolutionized banking practices and transformed the framework
of credit risk management. Credit risk modeling, valuation and rating have,
in consequence, been extensively studied. Many issues remain controversial
for both academics and industry practitioners, however, especially in the
wake of the recent financial crisis.
This thesis contributes to the existing literature in four ways. First,
a new approach is introduced for jointly estimating rating-specific bond
spread term structures. The approach achieves fitting accuracy comparable
to the more standard extended Nelson-Siegel technique meanwhile yielding
well-behaved and stable term structures which are appropriately ordered for
different rating classes.
Second, a ‘clinical’ study is presented of the evolution of Asset Backed
Security (ABS) valuations conditional on ratings during two crisis periods,
namely the 2007-2009 crisis and the earlier collapse of ABS secured against
Manufactured Housing loans of 2002-2003. The study focusses on the market’s
reactions to different pricing factors in crisis time including liquidity
and risk premiums.
Third, we examine predictable fluctuations in agency ratings behavior
over the business cycle and propose simple, tractable ways of parameterizing
time-varying rating transition matrices.
Fourth, we examine the dependence of corporate bond market risk premiums
on industrial sector. The analysis sheds light on the hypothesis that
the cross-sectional pattern of risk premiums reflects the degree to which
bond issuers operate in cyclically sensitive industries. By adjusting sectorspecific
spreads for time-varying and sector specific expected losses and for
liquidity, we show that spreads and, even more so, risk premiums are related
to the cyclicality of the obligor’s industry.
Date Issued
2010-11
Date Awarded
2011-01
Copyright Statement
Attribution NoDerivatives 4.0 International Licence (CC BY-ND)
Advisor
Perraudin, William
Creator
Wu, Shi
Publisher Department
Business School
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)