In this paper we focus on modeling and predicting the loss distribution for credit risky assets
such as bonds and loans. We model the probability of default and the recovery rate given default
based on shared covariates. We develop a new class of default models that explicitly account for
sector specific and regime dependent unobservable heterogeneity in firm characteristics. Based
on the analysis of a large default and recovery data set over the horizon 1980-2008, we document
that the specification of the default model has a major impact on the predicted loss distribution,
while the specification of the recovery model is less important. In particular, we find evidence
that industry factors and regime dynamics affect the performance of default models, implying
that the appropriate choice of default models for loss prediction will depend on the credit cycle and on portfolio characteristics. Finally, we show that default probabilities and recovery rates predicted out-of-sample are negatively correlated, and that the magnitude of the correlation
varies with seniority class, industry, and credit cycle.
Observable covariates are useful for predicting default, but several findings question their
value for explaining credit spreads. We introduce a discrete time no-arbitrage model with
observable covariates, which allows for a closed form solution for the value of credit default
swaps (CDS). The default intensity is a quadratic function of the covariates, specified such
that it is always positive. The model yields economically sensible results in terms of fit and the
economic impact of the covariates. Macroeconomic and firm-specific information can explain
most of the variation in CDS spreads over time and across firms, even with a parsimonious
specification. These findings resolve the existing disconnect in the literature regarding the
value of observable covariates for credit risk pricing and default prediction. Our results also
suggest that although CDS spreads are highly auto-correlated, analyzing spread levels may be
preferable to analyzing di¤erences for daily CDS data.
There are many unresolved issues in the modeling and calibration of credit risky
instruments that directly affect pricing and risk management: from the modeling of the
determinants of recovery rates, credit spreads, contagion, default dependence, to the testing of
models. In this paper we describe the current state of understanding regarding these different
areas and identify some of the unresolved issues.
In this paper we describe a methodology for deriving the upper and lower profit and loss (P&L) bounds in the presence of counterparty risk that does not rely on either structural or reduced form credit models. The methodology provides practitioners and regulators with a practical tool to estimate the impact on P&L of the two facets of counterparty risk: failure to perform and mark-to-market exposure. We show that for many applications, the bounds are tight and the credit worthiness of counterparties can have a major impact on the P&L.
This paper examines the different factors that have contributed to the subprime mortgage credit crisis: the search for yield enhancement, investment management, agency problems, lax underwriting standards, rating agency incentive problems, poor risk management by financial institutions, the lack of market transparency, the limitation of extant valuation models, the complexity of financial instruments, and the failure of regulators to understand the implications of the changing environment for the financial system. The paper sorts through these different issues and offers recommendations to help avoid future crises.
Recent court decisions, starting with the State Street decision in 1998, allow business methods to be patentable and now give financial institutions the option to seek patent protection for financial innovations. This new patentability paradigm and the heterogeneity of characteristics associated with financial innovations, poses an immediate decision problem for senior management: what to patent. We present a parsimonious decision framework that answers this question. We show that for innovations with certain characteristics, it is optimal not to patent, even if the option of patenting and licensing is available. Our model emphasizes the role of embedded real options that arise from certain types of Önancial innovations. The model provides an explanation of observed patenting behavior of financial institutions and the success of a wide class of innovations, including swaps, credit derivatives, and pricing algorithms.