Note(s): This table shows the results of a test of Hypothesis 2 using OLS regressions. The dependent variable, AbsAFE, is the absolute value of the median analyst forecast error for the period of ASC 606 adoption. Column 1 includes a propensity-score matched sample, matched on the treatment effect, which is an indicator equal to one if ASC 606 had a material impact on the firm’s net income in the period of adoption. The matching is done for all covariates in the analysis. Column 2 does not use a PSM sample and instead uses a continuous treatment variable, ASC606ImpactMagnitude. Each regression includes a sample of treatment and control firms for five quarters: Q1 2017–Q1 2018. Post = 1 for Q1 2018, the period of adoption of ASC 606. Each specification includes SIC 2-digit industry fixed effects. Variables are defined in Appendix A. Robust t-statistics are in parenthesis, with statistical significance denoted as ***p < 0.01, **p < 0.05, *p < 0.1