Profitability & returns
What it measures¶
This area asks whether the business earns a return on what it owns, and whether that return is getting better. A return here is a profit divided by the money tied up in the business: 6 of profit on 100 of assets is a 6% return on assets. The area runs five checks on the company's latest fiscal year:
| Check | The question | Basis |
|---|---|---|
| ROA positive | Was net income, divided by the assets at the start of the year, above zero? | absolute |
| CFO positive | Was the cash from running the business, divided by those assets, above zero? | absolute |
| ROA improving | Was the return on assets higher than the year before? | own history |
| Gross profitability | Is gross profit per unit of assets at least the sector's typical figure and the company's own five-year median? | sector band |
| Value creation | Does the return on the capital invested beat what that capital costs? | sector band |
Why it is on the page¶
A business that does not earn more than it spends cannot keep going without new money from lenders or owners, and one that earns less than its capital costs loses value as it grows. Studies of thousands of companies have found that profitability is the most persistent of the financial traits a company shows from year to year — which is why it is the first area of the band.
Thresholds, their source and the population they were fitted on¶
The area turns its checks into one figure: checks passed ÷ checks that could run, shown when at least three checks could run, and mapped to a verdict word by the guide's table.
Three of the checks come from Piotroski (2000), Joseph Piotroski's study of nine yes-or-no accounting checks on US companies whose shares were priced low against the equity on their balance sheet (a high book-to-market ratio). The sector figures come from Damodaran — Aswath Damodaran, a professor at New York University who publishes industry averages every January.
| Check | Line | Source and fitted population | Evidence grade |
|---|---|---|---|
| ROA positive | above 0 | Piotroski (2000): US companies with a high book-to-market ratio, 1976–1996 | peer-reviewed and replicated |
| CFO positive | above 0 | Piotroski (2000), same population | peer-reviewed and replicated |
| ROA improving | higher than last year | Piotroski (2000), same population | peer-reviewed and replicated |
| Gross profitability | at least the sector median and the own five-year median | Novy-Marx (2013): US companies 1963–2010, financial companies excluded; sector medians from Damodaran's January tables | peer-reviewed and replicated / practitioner convention |
| Value creation | return on invested capital at least the cost of capital | Koller and colleagues (McKinsey's valuation textbook); Damodaran's sector cost of capital | practitioner convention |
Accruals — the non-cash part of profit — are judged once, in Earnings quality & red flags, and never counted again here.
A worked example¶
Honeywell International, the US industrial group, on the statements stored on 25 October 2026 (fiscal year 2025):
- ROA positive — passed. Net income of 4,729 million US dollars on assets of 75,196 million at the start of the year: a return of 6.3%, above zero.
- CFO positive — passed. Operating cash flow came to 8.5% of the same assets.
- ROA improving — failed. The return on assets was 3.0 percentage points lower than in fiscal 2024.
- Gross profitability — not evaluable. Gross profit was 18.8% of assets, just under the company's own five-year median of 19.1%, but no sector benchmark is stored yet, so the check cannot judge its sector-band state.
- Value creation — not evaluable, for the same reason: the cost of capital comes from the sector benchmark.
Three checks could run and two passed: 2 of 3 checks passed · 67% · Good ✓, with the reason ROA positive (6.3%) meets > 0, absolute. The five-year history reads Strong, Good, Strong, Strong, Good from fiscal 2021 to 2025.
What “No verdict” means here¶
The tile shows No verdict when fewer than three of the five checks could run. That happens most often when the sector benchmark is missing (the last two checks need it) and an early fiscal year is missing too, so that the change check has nothing to compare with. The tile names the missing input and prints n of 5 checks could run.
Profitability is one of the four core areas: when it has no verdict, the band's overall figure is withheld as well. For a bank, an insurer or a holding company the tile reads No verdict — bank checks not yet available (or insurer, or holding company), because these checks do not describe their accounts.
What the verdict is and is not¶
The verdict describes how many of these published lines the company's latest statements met. Good says most checks that could run were met; it does not say the company is more profitable than its competitors (only the sector-band checks compare it with them), it is not a forecast of next year's profit, and it is not an instruction to trade the shares.