Sheen Liu

Dissecting Corporate Bond and CDS Spreads

In this article, the authors propose a new method to estimate the components of corporate bond and CDS spreads. They develop a CDS pricing model with default and nondefault factors and a corporate bond pricing model with default, tax, and liquidity factors using the reduced-form approach, and they jointly estimate parameters of both models from the pooled data. By formulating default intensity as a common factor in the prices of the CDS and reference bonds, the authors are able to identify the default and nondefault components of yield spreads more precisely. They find that, on average, the liquidity premium accounts for about 20–25% of corporate yield spreads across ratings and the size of the liquidity premium increases as the rating decreases. Furthermore, they find that the CDS spread contains a nontrivial nondefault component. Ignoring this component in the CDS spread thus results in a serious bias in the estimate of spread components when using the CDS information to aid in decomposition of corporate yield spreads.

Taxes, Default Risk, and Credit Spreads

Most term structure models of defaultable bonds have ignored personal tax effects. This research examines the effects of omitting taxes on empirical estimation of yield spreads and default probability. The model accounts for the effects of taxes on investors trading strategy. It is found that ignoring the interactive effects of taxes and default results in a substantial underestimation of yield spreads between corporate and Treasury bonds. The problem of spread underestimation is more severe in percentage terms for high-grade bonds than for low-grade bonds. Failure to account for tax effects results in an upward-biased estimate of default probability, a bias positively related to maturity and tax rates and negatively related to bond quality.

Effects of Credit Quality on Tax-Exempt and Taxable Yields

An examination of the effects of credit quality on the relative yields of Treasuries and municipals reveals that the municipal term structure model with no default risk underestimates equilibrium marginal income tax rates and yields of municipals relative to yields of Treasuries. The downward bias is more severe for long-maturity and low-rated bonds. Incorporating default risk into the municipal term structure model helps explain a considerable portion of the relative yields of tax-exempt and taxable bonds, and generates much more accurate estimates of implicit marginal income tax rates. After controlling for the effect of default, the estimated implicit tax rates for long and short bonds become quite comparable and consistent with the tax regime.

Inferring Default Correlation from Equity Return Correlation

This paper presents a new approach for estimating default correlation by linking default correlation to equity return correlation while preserving the fundamental relation between default and asset correlations in the structural framework. Our hybrid model thus overcomes a long‐standing empirical difficulty that default correlation estimation relies on the unobservable asset process. The empirical analysis shows that our hybrid model demonstrates a considerable improvement over the existing structural model of Zhou (2001) for the sample periods of 1970‐1993 and 1990‐2010. We also illustrate the difference between the two models in predicting default correlations over the period of the 2008 financial crisis.

Nondefault Components of Corporate Yield Spreads: Taxes or Liquidity?

Using the information in credit default swaps, we directly estimate the default, tax and liquidity components in the corporate yield spread from a generalized pricing model with taxes. Incorporating the credit default swap information facilitates disentangling the liquidity and default premia while including the tax rate in the pricing model helps identify the magnitude of tax premium. The model permits a more precise estimate of the liquidity premium by separating it from the tax component in the nondefault spread. We find that a substantial portion of the nondefault component in the corporate yield spread is due to taxes. The estimated liquidity component of the spread is highly correlated with bond-specific and marketwide liquidity proxies whereas the tax component is insensitive to these liquidity measures. Results indicate that the generalized model with taxes does a better job than the existing models in identifying the components of corporate yield spreads.

A Structural Approach for Predicting Default Correlation

Default correlation is a critical concept in risk management for fixed income investment, bank management, and insurance industry, working capital management, among many. We extend the Leland-Toft term structure model into a two-firm environment and predict the default correlation between two firms by directly simulating the calibrate model based on the observed equity data (1990-2010) for various ratings. Using our empirical default correlation estimation as the benchmark, our investigation sheds more light on the structural approach in predicting default correlation. The results show that our approach outperforms the previous Zhou’s model, thereby our approach is not only theoretical improvement, but also has an empirical advantage.