We revisit a method used by Das et al. (2007) (DDKS) who jointly test and reject a specification of firm default intensities and the doubly stochastic assumption in intensity models of default. The method relies on a time change result for counting processes. With an almost identical set of default histories recorded by Moody’s in the period from 1982 to 2006, but using a different specification of the default intensity, we cannot reject the tests based on time change used in DDKS. We then note that the method proposed by DDKS is mainly a misspecification test in that it has very limited power in detecting violations of the doubly stochastic assumption. For example, it will not detect contagion which spreads through the explanatory variables “covariates” that determine the default intensities of individual firms. Therefore, we perform a different test using a Hawkes process alternative to see if firm-specific variables are affected by occurrences of defaults, but find no evidence of default contagion.

We show that corporate bond issuers benefit from utilizing existing underwriter relationships when rolling over bonds, but at the same time become exposed to underwriter distress. A strong relationship enables the underwriter to credibly certify the issuer resulting in lower direct issuance costs and lower underpricing. However, if the underwriter becomes distressed, this spills over to the issuer's credit risk, because it weakens the relationship and increases the risk of involuntary relationship termination. The credit risk spillover is more pronounced for risky, information-sensitive issuers with high rollover exposure, i.e., those issuers most in need of certification by an underwriter.

Using a novel dataset of leveraged loan trades executed by managers of collateralized loan obligations (CLOs), we document the importance of “active loan trades” – trades executed at a manager’s discretion. More active trading increases the returns to CLO equity investors, lowers collateral portfolio default rates, and increases the manager’s chances of closing a new deal. Examining the observed loan trades, we find that more active CLOs trade at better prices than less active CLOs, selling leveraged loans earlier and before they get downgraded. Our findings suggest that more active CLOs are better at anticipating deteriorations in loan credit quality.