Return on Equity (ROE) Explained

Profit per Dollar of Owners' Capital

Return on equity divides net income by shareholders' equity: the profit the company earned for the year, measured against the capital its owners have tied up in it. A company earning $200 million of net income on $1.25 billion of shareholders' equity runs an ROE of 200 / 1,250 = 16 percent. Every dollar the owners have in the business produced sixteen cents of profit this year.

Shareholders' equity, the denominator, is the balance sheet's bottom line: total assets minus total liabilities. It represents the capital shareholders originally put in plus every dollar of profit retained since, minus buybacks and dividends paid out. Because equity moves during the year, careful calculations use the average of beginning and ending equity rather than a single snapshot.

The reason investors watch ROE is compounding. A business that can keep reinvesting retained profits at a 16 percent return grows intrinsic value far faster than one reinvesting at 6 percent, even if both look similar on this year's income statement. Sustained high ROE with room to reinvest is close to a working definition of a great business - which is exactly why the number attracts so much manipulation, deliberate and structural.

What Good Looks Like, Industry by Industry

The long-run average ROE for large US companies has hovered in the 12-15 percent range, which makes the mid-teens a reasonable first anchor. Persistently above 20 percent suggests either a genuine competitive advantage or an aggressive balance sheet; persistently below 8 percent suggests capital is compounding slower than shareholders could likely get elsewhere.

Industry context moves the goalposts substantially. Banks and insurers run enormous asset bases against thin equity by design, and a well-run bank posting 11-13 percent ROE is doing its job. Consumer brands with pricing power - think branded food, beverages, cosmetics - routinely post 20-30 percent because they need little capital to grow. Utilities cluster near their regulated allowed returns, typically 9-10 percent. Comparing a utility's 9 percent against a cosmetics company's 28 percent tells you about industry structure, not management skill.

The more informative comparison is the same company against its own history and its direct competitors. An industrial firm that has earned 14-16 percent ROE for a decade while peers earn 9 percent is demonstrating a durable edge. A firm whose ROE jumped from 12 to 22 percent in two years has a question to answer, and the balance sheet usually answers it.

The Leverage Trap, with Numbers

ROE has a structural blind spot: debt shrinks the denominator and can inflate the ratio without the business improving at all. Two companies with identical operations can post wildly different ROEs purely because of how they are financed.

Work it through. Company A runs $1 billion of assets funded entirely by equity. It earns $100 million of operating profit, pays 20 percent tax, and nets $80 million. ROE: 80 / 1,000 = 8 percent. Company B runs the identical $1 billion of assets and earns the identical $100 million of operating profit, but funds itself with $600 million of debt at 5 percent interest and only $400 million of equity. Interest costs $30 million, leaving $70 million pre-tax and $56 million after tax. ROE: 56 / 400 = 14 percent.

Same factories, same customers, same operating profit - yet B's ROE is nearly double A's. Nothing about B is better; B has simply substituted debt for equity, and B's shareholders now carry the risk that comes with it. If operating profit falls 40 percent to $60 million, A still nets $48 million for a 4.8 percent ROE, while B pays its fixed $30 million of interest, nets $24 million, and watches ROE drop to 6 percent - a much harder fall. Leverage amplifies ROE in both directions; the flattering direction is just the one that shows up in good years, which is when screens get run.

The defense is one extra glance: check the debt-to-equity ratio alongside ROE, or compare ROE against return on assets (ROA), which uses total assets as the denominator and cannot be flattered by financing. A high ROE sitting on a low ROA is leverage talking.

DuPont: Splitting ROE into Three Levers

The DuPont decomposition breaks ROE into three multiplied factors: profit margin (net income / revenue), asset turnover (revenue / assets), and financial leverage (assets / equity). In plain words: how much profit each dollar of sales keeps, how many dollars of sales each dollar of assets generates, and how many dollars of assets each dollar of equity supports. Multiply the three and the revenue and asset terms cancel, leaving exactly net income over equity.

Apply it to Company B from the leverage example, assuming $500 million of revenue. Margin: 56 / 500 = 11.2 percent. Turnover: 500 / 1,000 = 0.5. Leverage: 1,000 / 400 = 2.5. Multiply: 0.112 x 0.5 x 2.5 = 14 percent - the same ROE, now with its anatomy exposed. A luxury brand gets to a high ROE through fat margins and low turnover; a discount grocer gets there through thin margins and furious turnover; Company B gets there substantially through the 2.5x leverage factor.

The decomposition is most useful for explaining change. If a company's ROE rose from 14 to 18 percent, DuPont tells you whether margins expanded (often durable), assets started working harder (often durable), or leverage increased (a financing choice, not an operating improvement). Two companies with identical 18 percent ROEs can be a fortress and a house of cards, and the three factors are how you tell them apart.

When the Denominator Stops Making Sense

Buybacks distort ROE the same way debt does, by shrinking equity. Take the original company earning $200 million on $1.25 billion of equity - 16 percent ROE. If it spends $500 million repurchasing its own shares, equity drops to $750 million, and the same $200 million of profit now computes to 26.7 percent. Reported ROE jumped ten points while the underlying business did not change. For companies that have repurchased stock for decades, equity can shrink so far that ROE climbs into the hundreds of percent, at which point the ratio is describing balance-sheet history rather than business quality.

Equity can even go negative. Several large, healthy companies - Home Depot and McDonald's among them in recent years - have reported negative shareholders' equity after years of buybacks exceeding retained profits. Their ROE is literally uncomputable, yet the businesses are fine. The opposite case is uglier: big write-downs of failed acquisitions also shrink equity, so a company can post a rising ROE as a direct reward for having destroyed capital. Punishing the denominator is not a business achievement.

The honest workflow treats ROE as a starting point with two follow-ups: check what leverage and buybacks have done to the denominator, and check whether the return is sustainable by looking at its DuPont components over 5-10 years. Tracking how equity itself has compounded - a CAGR calculation on book value per share, alongside earnings growth - often says more about long-run value creation than any single-year ratio, and a DCF model translates those returns into what the stream of future profits is actually worth today.