Bond ratings and inter-rater agreement: a cross-sectional analysis