We occasionally select for retrospective review an all-time “best selling” research paper of the past few years from the General Financial Markets category of the Social Science Research Network (SSRN). In his January 2002 paper entitled “Value Investing: The Use of Historical Financial Statement Information to Separate Winners from Losers”, Joseph Piotroski applies a simple accounting-based fundamental analysis strategy to a broad portfolio of high (top 20%) book-to-market firms to enhance returns. His stock scoring system (FSCORE) consists of nine binary signals based on profitability and value-specific financial measures (see the list below). Using stock returns and fundamentals for a broad sample of U.S. stocks during 1976 through 1996, he finds that:
- Screening on accounting fundamentals boosts the average return earned by a high book-to-market investor by at least 7.5% annually.
- Extending this strategy to buy expected winners and short expected losers would have generated a 23% annual return between 1976 and 1996.
- The benefits of financial statement analysis are concentrated in small and medium-sized firms, companies with low trading volume and firms with no analyst following. Results suggest that the market is slow to react to historical financial information for such companies.
- One sixth of the difference in annual returns between strong and weak firms occurs during the four three-day periods around quarterly earnings announcements.
The author differentiates high book-to-market companies using signals based on nine financial parameters:
- ROA – net income before extraordinary items scaled by beginning of the year total assets
- CFO – cash flow from operations scaled by beginning of the year total assets
- Change in ROA – current year’s ROA less the prior year’s ROA
- Accrual – ROA-CFO
- Change in leverage – historical change in the ratio of total long-term debt to average total assets
- Change in liquidity – historical change in the firm’s current ratio (ratio of current assets to current liabilities at fiscal year-end) between the current and prior year
- Equity offering – whether the firm issued common equity in the preceding year
- Change in margin – current gross margin ratio (gross margin scaled by total sales) less the prior year’s gross margin ratio
- Change in turns – current year asset turnover ratio (total sales scaled by beginning of the year total assets) less the prior year’s asset turnover ratio
Each of the nine signals yields either a good (1) or bad (0) indication. The FSCORE for a company in a given year therefore ranges from 0 to 9. About 10% of the 14,000+ company observations earned a total score of 8 or 9; about 3% earned a score of 0 or 1.
In summary, careful screening of value stocks using accounting fundamentals may be able to enhance returns significantly.
Cautions regarding findings include:
- All return calculations are gross, not net. Including reasonable trading frictions, which are high during much of the sample period, would reduce reported returns. Outperformance appears to concentrate in the stocks most costly to enter and exit.
- Results also ignore data collection and processing costs (or management fee, if delegated).
Compare and contrast this approach with that of Partha Mohanram for growth stocks in “Mohanram’s Efficient Growth Investing”.
Frequently Asked Questions
What is Piotroski's FSCORE and how is it used in value investing?
Piotroski's FSCORE is a stock scoring system that sums nine binary accounting signals, each scored good (1) or bad (0), to produce a score from 0 to 9. It is applied to high book-to-market (value) stocks to separate likely winners from likely losers based on profitability, leverage, liquidity, and operating efficiency measures. In the study, about 10% of the 14,000+ firm-year observations scored 8 or 9, and about 3% scored 0 or 1.
Does screening high book-to-market stocks with accounting fundamentals improve returns?
Screening high book-to-market stocks with accounting fundamentals via Piotroski's FSCORE boosted the average return of a value investor by at least 7.5% annually over 1976-1996 in U.S. stocks. A long-short strategy buying high-scoring stocks and shorting low-scoring stocks would have generated a 23% annual return over the same period. These figures are gross returns and do not include trading costs.
How is the Piotroski FSCORE calculated?
The Piotroski FSCORE is calculated by assigning each of nine accounting signals a value of 1 (good) or 0 (bad) and summing them for a total of 0 to 9. The signals are ROA, cash flow from operations (CFO), change in ROA, accruals (ROA minus CFO), change in leverage, change in liquidity (current ratio), equity offering in the prior year, change in gross margin, and change in asset turnover. The study applied the score to the top 20% of firms by book-to-market using U.S. stock returns and fundamentals from 1976 through 1996.
Which types of stocks benefit most from Piotroski's financial statement analysis, and why might that be?
Piotroski's financial statement analysis benefits are concentrated in small and medium-sized firms, companies with low trading volume, and firms with no analyst following. The results suggest the market is slow to react to historical financial information for such companies. About one sixth of the annual return difference between strong and weak firms occurs during the four three-day windows around quarterly earnings announcements.
What are the limitations of the Piotroski FSCORE strategy's reported returns?
The Piotroski FSCORE strategy's reported returns are gross, not net, so reasonable trading frictions, which were high during much of the 1976-1996 sample, would reduce them. Outperformance also appears concentrated in the stocks that are most costly to enter and exit. The results further ignore data collection and processing costs, or management fees if the strategy is delegated.