Can a deep dive into company accounting data reliably predict stocks that will underperform? In their February 2019 paper entitled “Earnings Quality on the Street”, Urooj Khan, Venkat Peddireddy and Shivaram Rajgopal examine proprietary reports from a research firm that screens publicly available accounting data for over 9,000 North American and 5,500 other global companies to identify those in poor financial health (book value and earnings quality). Mutual funds, money managers, hedge funds, insurance companies, banks, CPA firms, law firms and individual investors subscribe to these reports. The research firm emphasizes importance of industry-specific metrics, evaluating whether each metric for a company is abnormal (aggressively optimistic) relative to peers and to its own history. Using 1,029 reports on aggressive reporting practices for 348 unique companies, and associated future daily stock returns, during 2004 through 2015, they find that:
- The most common accounting red flags relate to:
- Sales quality, especially managing revenues through customer financing, and other aggressive revenue recognition choices (65% of reports).
- Quality of margins (42% of reports).
- Flagged companies tend to be large and growing, with liquid stocks. They tend to score well on earnings quality models from prior academic papers and to have abnormal (very high or very low) accruals.
- On average, stocks of companies flagged for the first time have market-adjusted returns of -1.3%, -4.8%, -10.0%, -15.3% and -20.9% during the two trading days, three months, six months, nine months and one year after report publication, respectively. Corresponding 4-factor model (adjusting for market, size, book-to-market and momentum factors) alphas are -1.3%, -3.6%, -7.6%, -12.5% and -18.5%.
- Red flags are associated with future restatements, SEC Accounting and Auditing Enforcement Releases and GAAP-related lawsuits after controlling for other earnings quality indicators.
- A new earnings quality indicator for companies in retail, durable manufacturing and business services sectors based on rolling 5-year accounting data capturing 20 red flags from research firm reports adds value to other earnings quality metrics in predicting stock returns.
In summary, evidence suggests that deep analysis of company fundamentals can discover stocks that will substantially underperform the market over the next year.
Cautions regarding findings include:
- Returns are gross, not net. Costs of trading and shorting would reduce returns, as would the cost of subscribing to research reports. For long-only investors, the benefit of excluding flagged stocks from portfolios is unquantified.
- The paper does not identify the research firm.
- Construction of the new earnings quality indicator is beyond the reach of most investors, who would bear fees for delegating the work to an advisor. This new indicator is complex enough that its construction may impound material data snooping bias, thereby overstating its expected efficacy.
