ROE: the main measure of business quality and its main trap

ROE is the ratio of net profit to shareholders' equity: how much profit a business squeezes out of the money shareholders have left inside the company. The measure answers a question that neither revenue nor profit itself asks: how efficiently is the capital already invested being put to work? The main trap, though, sits in the denominator. Equity can be reduced — by buying back shares, by taking on debt instead of issuing new stock, by accumulating losses from previous years, by writing off a revaluation — and ROE will rise even though the business has not become the slightest bit better. A high ROE on its own does not distinguish a company that earns a lot from a company that simply has little equity.

What exactly goes into the numerator and the denominator

The numerator is net profit for the reporting period, that is, the figure after interest, taxes and non-recurring items. The denominator is shareholders' equity: share capital, retained earnings, reserves and revaluation, less the company's own shares bought back from the market. Two practical requirements follow directly from this. Profit is a flow over a period, while equity is a balance at a date, so it is more accurate to take average equity between the start and the end of the period rather than the closing figure: otherwise a large dividend payment in December will mechanically push the ratio up. And profit should be taken over a trailing twelve-month window, otherwise a seasonal quarter will distort the picture. The original reporting forms and the composition of an issuer's equity are better viewed in the financial reports section than in second-hand retellings.

The DuPont model: where a high ROE comes from

ROE is useful not as a final verdict but as a value that can be broken down into factors. The DuPont model splits it into net profit margin (how much profit is left from each rouble of revenue), asset turnover (how much revenue each rouble of assets generates) and financial leverage (how many times assets exceed shareholders' equity). The decomposition turns one figure into a diagnosis.

A retailer with a thin margin and high turnover, a manufacturer with a fat margin and slow-moving assets, and a heavily indebted company with mediocre operating efficiency can all report a similar ROE — but these are different businesses with different degrees of resilience. Only the first two factors can be regarded as quality: they tell you how the company operates. The third tells you only how it is financed.

The trap: leverage passes itself off as quality

Leverage cuts both ways. As long as the interest rate on debt is below the return on assets, borrowed money amplifies the return on equity. When rates go up and refinancing falls due, the same leverage amplifies the loss — and ROE collapses faster than it would at a business with no debt. This is why banks have a high ROE by nature: leverage is built into the business model, and comparing a bank with an industrial company on this measure is meaningless. The instrument card, together with the financial statements, provides context that the ratio itself does not contain: {{instrument:SBER}}.

There are also accounting sources of a false ROE. A share buyback reduces equity directly — and improves earnings per share along the way, an effect that deserves to be examined separately: earnings per share improve without any improvement in the business through exactly the same denominator mechanics. Accumulated losses from previous years eat into equity, and a company emerging from a difficult period suddenly reports a brilliant ROE. A company that has acquired another at a large premium carries goodwill among its assets — until it is written off, equity is inflated and ROE is understated; after the write-off it jumps. Large dividend payments also reduce equity and therefore mechanically lift the ratio — the payment dates can be seen here: {{dividend_calendar|limit=5}}, with more detail in the dividends section.

ROE and price: why compare at all

ROE by itself does not tell you whether a stock is worth buying — it speaks about the business, not about the trade. The link to price runs through book value: other things being equal, the higher and more stable the return on equity, the larger the premium the market is willing to pay for that equity. Hence a rule for reading the numbers: a high ROE alongside a high P/BV is not a contradiction but the price the market has already set for quality; a low P/BV alongside a low ROE is usually not a discount but a fair valuation of a weak return. It makes sense to compare the figure with the same issuer's price multiples — for example, 3,77 — and with the alternative that carries no equity risk: a return on equity below the yield on OFZ government bonds means the business does not justify holding a stake in it.

How to read the measure in Russian practice

Statements prepared under RAS and under IFRS give different equity and different profit: at a holding company, the parent's standalone accounts may show an almost empty balance sheet and a stream of dividends from subsidiaries — an ROE calculated on those accounts is meaningless. Use the consolidated statements. Non-recurring items — the sale of an asset, a foreign-exchange revaluation, a write-off — make profit incomparable between periods; it is worth looking at the trend over several years rather than at a single outstanding year.

One more caveat that should be stated plainly: ready-made industry benchmarks for a "normal" ROE do not exist, and this article is not going to invent them. The measure is meaningful only in comparison — with the company's own history and with competitors in the same industry; a list of stocks for such a comparison is available in the equities section, and definitions of the factors are in the glossary.

A short checklist

Step 1: take profit for the trailing year and average shareholders' equity. Step 2: break the result down using the DuPont model and see which factor is driving the figure. Step 3: check whether equity has been shrinking because of buybacks, losses or payouts. Step 4: compare with the company's history and with the industry, not with an abstract norm. If a high ROE holds for years with moderate leverage and a stable margin, that is most likely what business quality looks like. If it has risen within a quarter, look for the cause in the denominator first.