Earnings quality on the street

We develop a composite firm-year earnings quality score (EQSCORE) that uses signals based on fundamental analysis. We obtain a proprietary data set of 613 reports about aggressive reporting practices over 2004-2009 for 230 unique firms from a research firm (RF). From these reports, we identify red flags of poor earnings quality relating to (1) sales quality, (2) margin quality, (3) cash flow quality, (4) corporate governance, (5) audit, and (6) others. We construct the EQSCORE using 51 signals employed by the RF and a novel approach that imitates the RF's process for discovering earnings quality. The EQSCORE outperforms existing composite models of earnings quality in identifying accounting and auditing enforcement releases and restated firm-years and predicts future stock returns. The corporate governance-related and audit-related red flags included in the EQSCORE complement accounting-based red flags and enhance the ability of the EQSCORE to identify firms with poor earnings quality.