Return on Equity

Concept Analysis

Definition & Origins

Return on equity (ROE) measures how much profit a business generates on each dollar of shareholders' equity — the capital owners have supplied through paid-in capital and retained earnings. For Buffett it is not one ratio among many; it is his primary test of managerial economic performance. A business that consistently earns a high return on equity without leaning on debt is, in his framework, a structurally superior business.

The metric entered the letters early. In the 1977 letter Buffett dismissed the conventional definition of "record" earnings — a new high in earnings per share — noting that even a dormant savings account produces steadily rising interest each year through compounding. Except for special cases, he wrote, "a more appropriate measure of managerial economic performance" is return on equity capital. Berkshire's own 19% operating earnings on beginning equity that year counted for more in his eyes than its 37% gain in per-share earnings, because beginning capital was up 24% — much of the EPS gain was simply arithmetic.

Two years later, in the 1979 letter, he sharpened the formulation into its canonical form: the primary test of managerial economic performance is the achievement of a high earnings rate on equity capital employed, without undue leverage or accounting gimmickry — not the achievement of consistent gains in earnings per share (see Own Words).

Core Ideas

Earnings per share is the wrong metric. EPS growth can be manufactured by retaining earnings at almost any rate of return, because adding to the equity base mechanically grows future earnings even when the return on the additional capital is mediocre. A business retaining $100 million a year at an 8% return grows EPS reliably — and destroys value relative to paying the capital out to owners who could deploy it better elsewhere. The return earned on each incremental dollar of capital is what matters, not the aggregate EPS trajectory.

High ROE without leverage is the signature of a genuine business advantage. Most businesses can boost ROE by loading up on debt — leverage amplifies returns on equity arithmetically. But leveraged ROE is fragile: it depends on the earnings stream holding up well enough to service the debt. A business earning 20% on equity with no debt is a fundamentally different asset from one earning 20% with heavy borrowings. Buffett always reads ROE alongside the balance sheet. The 1983 Goodwill appendix makes the standard concrete: See's Candies earned its 25% after tax on net tangible assets with conservative accounting and no financial leverage.

The return on incremental capital is the forward-looking metric. Historical ROE describes the past; the return on new capital deployed describes the future. A business that earned 20% on equity when small but can find only 10% opportunities as it grows is a declining-quality business even while reported earnings keep rising. The 1984 letter formalizes this as the retention test: earnings should be reinvested only when they can be expected to earn high returns, and paid out when low returns are the likely outcome of reinvestment.

Retained capital must earn its keep. The dollar test from the 1984 letter: over time, each dollar of earnings retained should create at least a dollar of market value for owners. If retained dollars create less, they belong in shareholders' pockets. Buffett adds the accountability corollary — if earnings have been unwisely retained, it is likely that managers have been unwisely retained too.

Practical Application

The See's Candies benchmark. The 1983 letter's Goodwill appendix supplies the canonical numbers. Blue Chip Stamps bought See's in early 1972 for $25 million; the business then had about $8 million of net tangible assets and was earning about $2 million after tax — roughly 25% on tangible capital, achieved with no leverage. By 1983 See's was earning $13 million after tax on about $20 million of net tangible assets — a 65% return — having required only modest reinvestment along the way. That is what a high-quality ROE looks like in practice: the return rises as the franchise strengthens, rather than being diluted by ever-larger capital requirements.

The acquisition checklist. High ROE is not a ratio Buffett consults only after the fact; it is written into his buying criteria. The 1987 letter's standing invitation to sellers lists what Berkshire is looking for, and criterion three is "businesses earning good returns on equity while employing little or no debt" — paired with demonstrated consistent earning power and an explicit lack of interest in projections or turnarounds.

The textile counter-example. Berkshire's original textile operations earned poor returns on capital even in reasonable years, and every dollar reinvested in modernization earned less than its cost. The ROE logic runs in both directions: it eventually forced the wind-down, because allocating further capital to a business that cannot exceed its cost of capital destroys value regardless of revenue, employment, or sentimental attachment.

Reading management through the ratio. Managers who produce steady EPS growth by retaining earnings at low returns are failing the primary test even while appearing to succeed. The honest report card separates two questions: what return is the existing equity base earning, and what return is each newly retained dollar earning?

Common Misconceptions

Misconception 1: High ROE always signals business quality. Financial and highly leveraged companies can report impressive ROEs that reflect balance-sheet risk rather than economics. A bank earning 20% on equity through ten-to-one leverage carries a very different risk profile from an industrial company earning 20% with no debt. The ratio is only meaningful read together with the leverage that produced it.

Misconception 2: Declining ROE always means deterioration. As an outstanding business scales, it often cannot reinvest all of its earnings at the old rate, and ROE declines even though the core franchise is intact. The correct response is to return the excess capital — not to force it into lower-return ventures merely to preserve a ratio.

Misconception 3: Improving ROE is always genuine progress. ROE can be raised mechanically by shrinking the equity base through buybacks or special payouts, with no change in earning power. Conversely, a business investing heavily to widen its advantage may depress near-term ROE while building far greater future returns.

Misconception 4: Asset-heavy businesses are safer because their assets are real. The 1983 appendix dismantles this intuition: asset-heavy businesses generally earn low rates of return — rates that often barely fund the inflationary needs of the existing business, with nothing left over for real growth, distributions to owners, or acquisitions. In an inflationary world, the businesses needing the least tangible capital per dollar of earnings are hurt the least.

Thought Evolution

Related Concepts

Case Companies