Theory, research design assumptions, and causal inferences

Abstract

Ferri, Zheng, and Zou test Fischer and Verrecchia's (2000) prediction that a reduction in investors’ uncertainty about managers’ financial reporting objectives leads to an increase in the valuation-relevance of earnings reports. They use mandatory CD&A disclosures as an arguably exogenous “shock” that provided investors with more precise information about managers’ contractual incentives and find that these enhanced disclosures increased the relation between firms’ unexpected earnings and stock returns. Using Ferri et al. as a backdrop, we discuss the implicit assumptions invoked in natural experimental research designs and the fundamental role of theory in drawing causal inferences from empirical evidence.