Return on equity is one number. It's also one of the most misread numbers in investing. A 40% ROE looks like a dream stock, until you realize it might just mean the company is drowning in debt.
The ROE formula itself is simple. Net income divided by shareholder equity. What that number actually tells you is a different story, and it depends entirely on what's driving it.
What is ROE and why it matters
Return on equity measures how efficiently a company turns shareholder money into profit. Every dollar shareholders put into a business, either through direct investment or retained earnings, is supposed to work. ROE tells you how hard it's working.
A company with a 20% ROE generates 20 cents of profit for every dollar of equity on its balance sheet. That's the pitch. Compare that to a company earning 5%, and the first one looks like the better business.
Sometimes it is. Sometimes the comparison is meaningless, because the two companies are funding that profit in completely different ways.
This is why ROE is a starting point, not a verdict. It tells you a company is generating returns on equity. It doesn't tell you whether those returns are built on strong operations or built on borrowed money.
The ROE formula, step by step
The calculation itself takes seconds once you have the two numbers.
Both figures sit in a company's quarterly or annual filings. Net income comes from the income statement, published every quarter alongside the rest of the earnings report, and the Stoxcraft Earnings Calendar is a fast way to see when the next one lands for a given name.
Shareholder equity comes from the balance sheet, and it's what's left over after every liability is subtracted from every asset. Some analysts use average equity across the period instead of the ending balance, which smooths out swings from buybacks or capital raises during the year. For a quick read, the ending balance works fine.
One thing worth flagging early. ROE only means something when both net income and equity are positive. A company with negative equity can show a wildly high or even positive ROE off a net loss, and the number becomes noise instead of signal.
Odyssey Marine Exploration (OMEX) is a real case in point: negative shareholder equity combined with a net loss currently produces a headline ROE north of 100%, a number that looks like a top-tier compounder and means nothing of the sort. If equity is negative, skip the ratio and look at the balance sheet directly.
The DuPont breakdown: when high ROE is a red flag
ROE can be pushed up two very different ways, and only one of them is a sign of a good business. Most explanations skip this part entirely.
The DuPont formula breaks ROE into three pieces: net margin, asset turnover, and financial leverage. Multiply the three together and the result is ROE. Net margin shows how much profit survives out of every dollar of revenue.
Asset turnover shows how efficiently assets generate sales. Financial leverage shows how many dollars of assets are backed by every dollar of equity, and it's the piece that debt directly inflates. What this breakdown reveals is that a company can post an excellent ROE purely by loading up on debt, shrinking its equity base, without actually running a more profitable business.
Innodata (INOD), a data-engineering company that preps training data for AI systems, shows what the good version of a high ROE looks like. Its ROE runs above 30%, and it gets there with almost no debt on the balance sheet at all. That's margin and efficiency doing the work, not leverage.
Willis Lease Finance (WLFC), an aircraft-engine leasing company, tells the opposite story at a lower headline number. Its ROE sits in the mid-teens, on paper a more modest figure than Innodata's. But Willis carries total debt several times larger than its equity base, the kind of leverage that's structurally normal for a leasing business but still means a meaningful share of that ROE is coming from borrowed capital, not from pure operating efficiency.
Willis also pays a small dividend, currently yielding under 1%, worth running through the Stoxcraft Dividend Calculator for anyone comparing income against the leverage story here. Every dollar paid out as that dividend is a dollar that leaves retained earnings, and by extension shareholder equity, instead of building it.
None of this makes leverage bad by definition. It just means a high ROE built on debt carries more risk than a high ROE built on margins, because that debt still has to be serviced regardless of how the business performs. Checking net margin and debt to equity alongside ROE takes a few extra minutes and tells you which story is actually in front of you.
The three-way comparison holds up on the chart as well as on the scorecard.
What counts as a "good" ROE by sector
There's no universal cutoff for a good ROE, and treating one like it exists is where a lot of comparisons go wrong.
Asset-light businesses like software companies naturally run high ROE, because they need relatively little equity capital to generate revenue. Capital-intensive industries like utilities, industrial manufacturers, or REITs carry large equity bases by nature, so their ROE tends to sit lower even when the underlying business is perfectly healthy.
A REIT running a 6% ROE isn't necessarily a worse business than a software company running 25%. They're playing different games with different capital structures.
This is why ROE only means something in context. The useful comparison is a company against its direct peers, not a company against the market average. If most of a sector clusters around 10 to 15%, a name sitting at 5% is worth a closer look.
A name sitting at 40% in that same sector deserves the DuPont check above before anyone calls it the better business.
Banks are their own special case. Their equity base is small relative to the size of their loan books by design, which pushes ROE structurally higher than in most other industries. A bank posting a 15% ROE isn't automatically outperforming a consumer staples company posting 18%.
They belong to different capital models, and stacking them side by side on ROE alone hides more than it reveals.
The fix is simple even if it takes a bit more digging: pull up two or three direct competitors before deciding whether a given ROE is actually good or just average dressed up as impressive. The Stoxcraft Screener is a fast way to pull that peer set without digging through filings one ticker at a time.
ROE vs ROA: what's the difference
Return on assets answers a related but distinct question. Instead of measuring profit against equity, ROA measures profit against total assets, debt and equity combined.
That difference matters more than it sounds. A highly leveraged company can show a strong ROE while its ROA stays mediocre, because ROA doesn't let debt-funded assets flatter the result the way equity-only math does.
Comparing the two side by side is a fast way to spot how much of a company's return is coming from leverage versus from actually running the business well. A large gap between a high ROE and a low ROA is usually the leverage story showing up again.
ROE tells you efficiency, not safety
A high return on equity is a starting point for a question, not an answer on its own. It says a company is generating strong returns on the equity shareholders have put in. It says nothing, by itself, about whether that return came from a genuinely efficient business or from a balance sheet stretched thin with debt.
Run the DuPont breakdown, check the ratio against real sector peers, and put ROA next to it before drawing a conclusion. Do that, and ROE stops being a number that can fool you and becomes one of the more useful shortcuts in fundamental analysis. For a deeper walkthrough of reading a company's full numbers, the Fundamental Analysis island covers the rest of the balance sheet in the same style.
