What’s my target? Analyst forecast dispersion and earnings management through effective tax rates

Abstract

Kirk, Reppenhagen, and Tucker (2014) report that, consistent with the existence of private information, investors use individual analyst forecasts as additional benchmarks to evaluate reported earnings. Following this logic, we investigate whether managers consider the private information in a subset of analyst forecasts when managing earnings. Specifically, we test whether changes in year-end tax accruals are associated with analyst forecast dispersion, our measure of private information. We find that when pre-managed earnings would have beat the consensus and analyst private information is low (i.e., dispersion is low), managers increase tax expense and create cookie jar reserves. When analyst forecasts reflect increased levels of private information (i.e., dispersion is high), we find that firms use tax expense to further increase earnings even when pre-managed earnings would have beat the consensus. Additional analyses reveal that the effect of dispersion is conditional on the proximity of pre-managed earnings to the consensus forecast. Our results highlight how managers consider individual analyst forecasts to calibrate earnings management and contribute to our understanding of earnings management activity around consensus estimates.