The impact of discretionary segment reporting behavior on investor beliefs and stock prices.

Abstract

This dissertation examines the determinants of and the informational and valuation consequences of discretionary segment reporting behavior. Using a simultaneous framework to control for effects of endogeneity, the first essay in this dissertation (1) identifies factors associated with variation in time-series and cross-sectional segment reporting choices and (2) investigates the effect of segment reporting choices on the precision of investor beliefs. Time-series evidence suggests that information asymmetry and poor relative valuation are factors associated with the decision to increase the number of segments reported, while cross-sectional tests document that segment reporting choices reflect a trade-off between proprietary cost-related factors and valuation-based benefits of disclosure. Using two-stage generalized least squares to control for endogeneity, a discretionary expansion of segment reporting fineness is associated with a ten percent improvement in analyst forecast accuracy and a seven percent reduction in forecast dispersion. Cross-sectional tests show that segment reporting choices have a positive (negative) influence on analyst forecast accuracy (dispersion) after controlling for other factors in the firm's information environment. The second essay examines whether a discretionary expansion of segment reporting communicates value-relevant information and, given the decision to improve disclosure, yields the positive valuation revisions hypothesized by existing disclosure models. Existing theoretical research shows that firms will improve disclosure when the benefits exceed the costs; therefore, firms that choose to improve reporting practices should experience positive valuation revisions. Consistent with extant theory, firms that choose to voluntarily expand the number of segments reported experience positive earnings forecast revisions and market-adjusted stock returns in the period surrounding the reporting increase. These positive returns reflect both the improvement in consolidated earnings expectations as well as the incremental earnings contribution of the newly reported segment. After controlling for this mean disclosure effect, the improvement in investor precision that accompanies these reporting changes also has a significant impact on stock prices. Finally, price-earnings regressions document that a dollar of expected earnings is being capitalized at a higher valuation multiple after the reporting change. The results suggest that firms with an incentive to expand disclosure garner the valuation benefits predicted by existing disclosure models.