Analysts’ role in shaping non-GAAP reporting: evidence from a natural experiment

We examine how exogenous changes in analyst coverage influence (1) the likelihood that managers will voluntarily disclose customized (non-GAAP) performance metrics and (2) the relative quality of their non-GAAP disclosures. Specifically, we use a quasi-natural-experimental setting in which brokerage firms terminate analyst coverage and find that, following an unanticipated decrease in analyst coverage, managers are more likely to disclose non-GAAP earnings per share (EPS) numbers. We also find that managers become more aggressive in their disclosure choices and that the quality of their non-GAAP exclusions decreases after analysts terminate coverage. These effects are more pronounced among firms losing an analyst with greater ability and firms with weaker corporate governance. Overall, our evidence suggests that analysts’ monitoring deters aggressive non-GAAP reporting.