The following entry appeared as part of Governance Metrics International’s GMI Blog. GMI is the leading independent provider of global corporate governance and ESG ratings and research. Corporate stakeholders – including leading investors, insurers, auditors, regulators and others – use GovernanceMetrics services to identify and monitor risks related to non-financial measures covering key environmental, social, governance and accounting risk factors.
Academic Research Validates Relevance of GMI Ratings’ Accounting & Governance Risk
Accounting & Governance Risk (AGR®) ratings express GMI Ratings’ assessment of the accuracy and reliability of a company’s financial reporting and the quality of its governance. AGR is based on metrics that have been identified as being the most associated with SEC enforcement actions for accounting fraud. In fact, a recent academic study shows that the AGR is superior to many academic risk measures for predicting accounting irregularities.
In Detecting and Predicting Accounting Irregularities: A Comparison of Commercial and Academic Risk Measures (Available here) authors Price, Sharp and Wood set out to compare a substantial body of academic research on developing risk proxies to detect accounting irregularities with a commercial measure developed for the same purpose by Audit Integrity (now GMI Ratings). The study is based on firms in the Compustat database from 1995 to 2008, using annual data.
Audit Integrity’s AGR measure was compared to six measures from academia. The study compared the ability of these measures to detect and to predict three different variables:
・SEC accounting and auditing enforcement releases (AAERs),
・restatements related to accounting irregularities, and
・shareholder litigation for accounting improprieties.
A head-to-head comparison of each academic measure relative to AGR was conducted. AGR outperformed the academic risk measures in all head-to-head tests for detecting, and in most head-to-head tests for predicting, all three of these sets of events.
The results showed that of the 36 comparisons between AGR and the six academic risk measures, AGR was superior to the academic measures in 28 of the comparisons. When the comparison was in terms of detecting accounting irregularities, AGR performed better in 18 out of the 18 comparisons. When the comparison was in terms of predicting irregularities one period before the irregularities would occur, AGR performed better than the academic measures in 10 out of the 18 tests.