A Parsimonious, Simple, Theory-Based Measure of Reporting Quality: Omissions of Seven Key Financial Statement Variables

• In Event: Accounting and Disclosure

Abstract

Chen et al. (2015) propose a measure of disclosure quality, DQ, based on the frequency of missing financial statement variables; computed with hundreds of items and a complex measure to program. They validate their model based on its associations with four accepted, extant measures of disclosure quality: analyst forecast accuracy and dispersion, bid-ask spread, and cost of capital. We propose a simple, parsimonious, theory-based and empirically-supported measure of missing variables, REPORT, based on the frequency of omissions of the seven variables found most value relevant by Lev and Thiagarjan (1993). To compare REPORT with DQ, first, we replicate Chen, et al. (2015) with a current sample and find results for DQ as they originally reported. REPORT is similarly associated with these validation metrics. Thus, overall, we conclude that the two constructs perform similarly well. Second, with Vuong and Clarke tests we directly compare the relative strength of REPORT’s and DQ’s associations with these four measures and find that REPORT performs relatively very well compared to DQ; DQ is found superior to REPORT in only two of the eight tests. REPORT, being theory-based, omits examination of many items likely unnecessary to firm valuation and is easily implemented, not only by trained researchers, but also by average investors; a disclosure quality measure that is readily operationalized for both academic and applied purposes.

Authors

• David G Harris, Syracuse University

• YING ZHANG, Syracuse University