The Effect of Accounting Reporting Complexity on Financial Analysts

Abstract

We investigate the association between a new XBRL based measure of accounting reporting complexity (ARC) and analyst behavior. We find that analysts are less likely to cover firms with complex accounting. Further, higher ARC is associated with lower forecast accuracy, higher forecast dispersion, and lower informativeness of recommendation revisions and responsiveness to earnings announcements. This association is attenuated when analysts have longer tenure, greater firm-specific experience, and are focused on fewer industries. Investigating several complex accounts, we find that the complexity of derivatives, fair value, and pension accounts are each negatively associated with forecast accuracy, suggesting that understanding these complex accounts requires specialization. We propose a new measure of analysts’ account-specific expertise and find that expertise with derivative and fair value accounts attenuates the negative effects of complexity in these accounts to a greater extent than general analyst experience. Overall, our findings suggest that analysts’ expertise plays an important role in mitigating the adverse effects of ARC.