Have ratings become more accurate?
研究发现评级机构的方法改进使评级更准确，能更好预测违约，并与会计指标和市场风险指标更一致，解释了评级标准趋严的现象。
Prior studies indicate that rating agencies have adopted more stringent rating criteria over time. In this paper, we hypothesize that improvements in rating accuracy can explain some of these observed patterns. We present empirical evidence supporting this hypothesis, demonstrating that enhancements in rating methodologies have resulted in better default prediction. Our analysis also reveals that, over time, ratings have become more closely aligned with accounting fundamentals and a market-based measure of default risk (distance-to-default). These findings provide a fresh perspective on the factors influencing changes in credit rating standards and emphasize the significance of methodological advancements in credit risk assessment. This research introduces the novel argument that enhancing rating accuracy is an economic rationale for long-term rating trends. The findings underscore the continued importance of credit ratings despite criticisms, suggesting that ratings remain a valuable tool for investors.