The idea. Joseph Piotroski ("Value Investing: The Use of Historical Financial Statement Information to Separate Winners from Losers," Journal of Accounting Research, 2000) attacked a different problem. Cheap value stocks (highest book-to-market quintile) beat the market on average, but the average hides a graveyard. Most individual value stocks underperform, and a few recover spectacularly. Could nine simple, binary accounting signals, each worth one point with no weights and no regression, separate them in advance? They could. Within his 1976 to 1996 value universe, high-F firms beat the value portfolio's average one-year market-adjusted return by about 7.5 percentage points. A long-high-F, short-low-F strategy returned about 23% annualized (his figures, and the paper is free, so read it). The F-Score's genius is its deliberate crudeness: nine yes/no questions any analyst can answer from two years of statements in five minutes.
What are the nine Piotroski F-score signals?
The nine signals (score 1 if true, else 0). Piotroski's scaling conventions matter if you want to match his numbers. ROA and CFO are scaled by beginning-of-year total assets, and leverage uses average total assets.
Profitability:
- ROA > 0: net income before extraordinary items / beginning TA is positive.
- CFO > 0: operating cash flow is positive.
- ΔROA > 0: ROA improved vs prior year.
- ACCRUAL: CFO > ROA. Cash earnings exceed accrual earnings. It's Sloan's ratio as a binary, and since both are scaled by beginning TA, this is CFO > net income.
Leverage, liquidity, and dilution:
- ΔLEVER < 0: long-term debt / average TA fell.
- ΔLIQUID > 0: current ratio rose.
- EQ_OFFER = no issuance: the firm did not issue common equity during the year. A cheap company selling equity is confessing it can't fund itself.
Operating efficiency:
- ΔMARGIN > 0: gross margin ratio improved.
- ΔTURN > 0: asset turnover (sales / beginning TA) improved.
Advertisement
How do you read the Piotroski F-score?
Reading the total: F ≥ 8 = strong, with fundamentals improving on all fronts (Piotroski's buy zone within value). F ≤ 2 to 3 = weak: deteriorating, his short zone. 4 to 7 is the muddled middle. Two design notes. The score was built for value stocks, so inside a glamour-growth universe its signals mean less, since a hypergrowth firm rationally issues equity and lets margins dip. And every signal is a change, so F measures direction of travel and says little about the level of quality. A mediocre business getting better outscores a great business having a soft year. Use it that way. F is the screen for "is this cheap company fixing itself?", which is the question a value analyst needs answered before the falling knife hits the floor.
Worked example: scoring a company on the nine signals
Worked computation: Kavery. Beginning-of-Y1 TA = 900 (given). Signals for Year 2:
| # | Signal | Computation | Point |
|---|---|---|---|
| 1 | ROA > 0 | 130 / 1,000 = 13.0% > 0 | 1 |
| 2 | CFO > 0 | 25 > 0 | 1 |
| 3 | ΔROA > 0 | 13.0% vs 91/900 = 10.1% | 1 |
| 4 | CFO > ROA | 25/1,000 = 2.5% vs 13.0% (no) | 0 |
| 5 | ΔLEVER < 0 | 200/1,200 = 16.7% vs 150/950 = 15.8% (rose) | 0 |
| 6 | ΔLIQUID > 0 | 660/376 = 1.76 vs 450/250 = 1.80 (fell) | 0 |
| 7 | No equity issued | ₹94 crore QIP (issued) | 0 |
| 8 | ΔMARGIN > 0 | 38.0% vs 40.0% (fell) | 0 |
| 9 | ΔTURN > 0 | 1,300/1,000 = 1.30 vs 1,000/900 = 1.11 (rose) | 1 |
F = 4 / 9. Every screen has now told the same story: headline growth and profits (signals 1 to 3 score) sitting on top of deteriorating cash conversion, margins and liquidity, plus fresh dilution. One synthetic company, four instruments, one consistent diagnosis. Read the receivables, the soft assets and the QIP documents before you believe a single rupee of the growth.