Kelly Nianyun Cai

Fallen Angels and Price Pressure

We examine price pressure in a setting where trades occur because of regulations and when information effects are absent. Our study of fallen angel bond sales by insurance companies shows that price pressure is negligible, if not non-existent. We attribute our results to the fact that the trades occur when fundamentals are unchanged and dealers know that the sales are not motivated by private information about future returns. Our results confirm the predictions of Admati and Pfleiderer (1991) and Roell (1990) that sellers will benefit from a higher price when dealers recognize that they are uninformed. We find that insurers do not attempt to hide their trades by selling bonds before they are downgraded, consistent with following a strategy of sunshine trading.

Effect of Being a Fallen Angel on Bond Ratings and Yields

This article investigates whether bond ratings serve as a basis for market prices of bonds. Unlike most previous work, which focused on rating changes, the author uses a new sample to consider bond ratings at issue. The author compares the yield spread of general high-yield bonds and high-yield bonds issued by fallen angels (firms that no longer carry an investment-grade rating). If investors totally depend on ratings, equally rated bonds should have the same yield spreads. This study finds, however, that bonds of identical ratings have different yield spreads. Specifically, for the BB/Ba rating category, yields of high-yield bonds issued by fallen angels are higher than the yields of similarly rated bonds issued by firms that have a speculative rate from the beginning. These findings suggest that bond ratings do not perfectly capture information related to bond characteristics in general. Results also suggest that rating agencies are too optimistic when rating high-yield bonds issued by fallen angels.

Debt IPO Waves

We examine aggregate volume of straight debt IPOs issued by nonfinancial firms over an extended period of 1970 to 2010. We find that aggregate debt IPO activities display wave patterns. Similar to equity IPOs, both the number and total proceeds of debt IPOs vary substantially over time. We explore possible explanations for the debt IPO waves with four groups of variables - capital market conditions, investor sentiment, information asymmetry, and interest rates. Our results indicate that debt IPO volume is significantly associated with term spread, stock return volatility and interest rates, suggesting that bond market conditions and information asymmetry play significant roles in explaining time variations in debt IPO volume. However, we do not find a significant role for investor sentiment. We also document that debt IPOs and equity IPOs are mutually Granger caused, suggesting that debt IPOs and equity IPOs tend to move together.