Profitability ratios answer one question from several angles: how efficiently does a company turn its inputs — sales, capital, and owner equity — into profit? Margins measure profit per dollar of revenue (how much of each sale the firm keeps). Return on equity (ROE) measures profit per dollar of owner capital. Return on invested capital (ROIC) measures operating profit per dollar of all capital tied up in the business, debt and equity alike. Together they form the core of fundamental "quality" analysis. The defining tension of the whole cluster is that profitability is the most economically meaningful family of ratios and the most easily distorted — by leverage, by accounting choices, and by the human habit of extrapolating a few good years into a permanent moat. None of these numbers means much in isolation; their value comes from decomposition, trend, and peer context.

What this section covers

This is the section overview for four interlocking sub-topics. Read the child nodes for the mechanics and evidence; this page maps how they relate.

• Profit Margins (Gross / Operating / Net) — the income-statement "waterfall." Gross strips only direct cost of goods; operating also removes overhead, R&D, and selling costs; net removes interest, tax, and one-offs. Margins read the business per sale and localize where profit is made or lost.
• Return on Equity (ROE) — net income over shareholders' equity; the bottom-line return on owner capital. The most-cited quality number, and the most easily flattered by debt.
• DuPont Analysis — the diagnostic glue. It factors ROE into margin × asset turnover × equity multiplier (and a 5-step form isolating tax and interest), revealing why an ROE is what it is rather than just what it is.
• Return on Invested Capital (ROIC) — NOPAT over invested capital; capital-structure-neutral, so it isolates operating capital efficiency. Read against the cost of capital (WACC), it is the tightest accounting link to value creation.

The connective logic: margins feed DuPont (net margin is one of its three terms); DuPont explains ROE; and ROIC is the leverage-neutral cross-check on ROE — when DuPont shows ROE propped up by a rising equity multiplier, ROIC tells you whether the underlying business actually earns its cost of capital. An analyst rarely uses one without the others.

The core tension: real quality vs. flattered numbers

The single most important idea in this section is that the same ratio can signal opposite things. A high ROE can mean a durable franchise reinvesting at high rates — or a thin, debt-loaded equity base where leverage is doing the work. A high gross margin can reflect genuine pricing power — or cost misclassification (peer-reviewed "classification shifting" research documents managers shifting COGS into SG&A to inflate gross margin while net income is unchanged; see the Margins node). A glowing ROIC can mark a true compounder — or simply old, depreciated assets understating book capital, or the top of a mean-reverting fade.

This is why the section's recurring discipline is the same across all four nodes:

1. Decompose before you judge (DuPont for ROE; the gross→operating→net waterfall for margins; the ROIC–WACC spread for capital returns). 2. Trend over level — a rising component is usually a better signal than a high static one. 3. Peer- and industry-relative only — banks, utilities, software, and grocers have structurally different margins, ROEs, and ROICs; cross-industry comparison is the most common abuse.

When it matters — and when it doesn't

Profitability ratios matter most for assessing the durability and quality of a business over multi-year horizons: screening for moats, comparing business models, feeding the value-driver assumptions in a DCF, and screening out capital-destroying firms (ROIC persistently below WACC). They matter least as timing or valuation tools — they say nothing about whether a stock is cheap. The cluster is also near-useless for financial firms (banks, insurers), where "assets," "leverage," and "invested capital" mean something entirely different and DuPont in particular breaks down (Wikipedia; standard FSA texts).

Adoption, debate & evidence

These ratios are near-universal and uncontested as descriptive measures — taught in every CFA and corporate-finance curriculum, reported by every data vendor. The honest debate is about predictive power, and here look-alikes must be kept separate. The rigorous evidence for a "profitability premium" comes from the academic quality-factor literature: Novy-Marx (2013) showed gross profitability predicts the cross-section of stock returns about as well as value metrics, because gross profit is relatively unpolluted by accounting accruals; the profitability factor was important enough that Fama-French added it (along with investment) to their five-factor model, and it is the "RMW"/"ROE" leg of the q-factor models (Hou-Xue-Zhang). Related work (Ball et al.; Asness "Quality Minus Junk") finds ROA, ROE, gross margin, and cash-based profitability all carry signal.

Three caveats are essential and frequently lost:

• The premium belongs to the broad, long-short, cross-sectional factor, not to "buy the highest-ROE/ROIC stock." A great business is not automatically a great investment if quality is already priced in — returns come from the gap between fundamentals and expectations.
• Changes in profitability often predict better than levels (the q-factor ROE leg uses recent quarterly ROE; DuPont research finds the change in ROA is more informative than its level).
• Mean reversion is powerful. Mauboussin's long-run studies show ROICs broadly fade toward the cost of capital; genuine persistence exists but is the minority of firms. Treat "high profitability = buy" as unproven and price-dependent.

Popular thresholds ("ROE above 15–20% is good," etc.) are heuristics, not validated cutoffs — the right benchmark is always the peer group and the firm's own history.

Strengths & limitations (section-level)

Strengths. Directly tied to owner and operating economics; decomposable into transparent causal chains; among the more durable quality signals when read as multi-year trends versus close peers.

Limitations / shared failure modes. Leverage flatters ROE; book-value accounting distorts ROE and ROIC (old assets understate capital; unrecorded intangibles overstate returns for asset-light firms); net income and margins are gameable via one-offs, non-GAAP "adjustments," and cost classification; there is no canonical ROIC formula (goodwill, leases, cash, and averaging conventions vary by source). The #1 cross-cutting misuse: reading a single profitability number at face value — without decomposition, a debt check, an industry benchmark, and a trend — and especially comparing it across industries.

Sources

• Robert Novy-Marx (2013), "The Other Side of Value: The Gross Profitability Premium," JFE — gross profitability as a return predictor: https://mysimon.rochester.edu/novy-marx/research/OSoV.pdf
• Hou, Xue & Zhang — q-factor model (ROE/profitability factor), NBER w24709: https://www.nber.org/system/files/working_papers/w24709/w24709.pdf
• AlphaArchitect — "The Profitability Factor: International Evidence" (robustness of profitability metrics incl. ROE, ROA, gross margin): https://alphaarchitect.com/the-profitability-factor-international-evidence/
• NBIM Discussion Note 3-15 — "The Quality Factor" (QMJ components: ROA, ROE, GPOA, gross margin): https://www.nbim.no/contentassets/0660d8c611f94980ab0d33930cb2534e/nbim_discussionnotes_3-15.pdf
• Aswath Damodaran (NYU Stern) — return measures and margins by sector: https://pages.stern.nyu.edu/~adamodar/pdfiles/papers/returnmeasures.pdf
• Corporate Finance Institute — profitability ratios overview: https://corporatefinanceinstitute.com/resources/accounting/profitability-ratios/
• Child nodes (this section): DuPont Analysis, Margins, ROE, ROIC — see their own Sources for primary citations (Soliman 2008; Fairfield-Yohn 2001; Nissim-Penman 2001; Mauboussin; McKinsey Valuation).

Disputes flagged: the profitability premium is real but belongs to the academic long-short factor, not to single-stock "highest-ratio" screening; changes in profitability predict better than levels; ROIC has no canonical formula; popular ROE/margin thresholds are folklore heuristics; and whether currently elevated aggregate margins mean-revert is contested macro opinion.