Information Content of Analysts' Forecasts in the Bond Market

Abstract

Motivated by the paucity of studies examining analysts' forecasts in the bond market, this study addresses the question whether analysts' forecasts are informative to debt rating agencies and bondholders. Specifically, I investigate whether analysts' forecasts provide any incremental explanatory power to cost of borrowing. I construct simultaneous equations by using debt ratings and bond yield premium to measure the cost of borrowing. I apply three-stage-least-squares (3SLS) procedure incorporating firms' industry-adjusted earning forecasts, cash flow forecasts and growth forecasts into the model. Additionally, I examine whether the information content of these forecasts to bond investors vary in different contexts. The results show that analysts' forecasts capture information used by rating agencies to assign debt ratings, while overall bond investors rely heavily on credit ratings and analysts' forecasts do not provide significantly incremental explanatory power in explaining bond yield premium. However, when perceived default risk is high, or when bond investors perceive a high level of private information collection by analysts in the context of high information asymmetry or highly unpredictable earnings, analysts' forecasts become informative to bond investors as well. I also find that analysts' forecasts capture less information used by bond investors after regulation Fair Disclosure (FD) took effect and debt investors rely more heavily on debt ratings in the post-FD period.