ROE vs ROCE vs ROA: Which Ratio Matters?

ROE, ROCE and ROA are three "return ratios" that answer one question in slightly different ways: how good is this company at turning money into more profit? ROE measures the return on the owners' money, ROCE the return on all long-term capital including borrowed funds, and ROA the return on the entire asset base.

The trap is simple: the highest of the three numbers is often the most dangerous one. A flattering ROE can be genuine quality, or it can be borrowed money wearing a disguise — and you cannot tell which from that single number alone.

The relief is just as simple. Once you place ROE next to ROCE, the trick becomes visible in about thirty seconds. The gap between the three ratios is where most beginners lose the plot — and it is also where the answer hides.

What each ratio actually measures

Think of a business as a kitchen. Money goes in, profit comes out. The three ratios just measure the same kitchen using different yardsticks for what counts as the input.

Return on equity (ROE) is net profit divided by shareholders' equity — ideally the average equity across the year, since that balance keeps moving as profit is retained. It tells you how many rupees of profit the company makes for every rupee of the owners' money parked in the business. If a company has ₹100 of equity and earns ₹18 of profit, the ROE is 18 per cent. This is the number the owners care about most, because it is the return on their slice alone.

Return on capital employed (ROCE) is operating profit — EBIT, the profit before interest and tax — divided by capital employed, the long-term money in the business (total assets minus current liabilities, which works out to roughly equity plus debt). It ignores how the money was raised and just asks how well the business uses the whole pool. A company with ₹100 of equity and ₹100 of long-term debt earning ₹24 of EBIT has an ROCE of 12 per cent.

Return on assets (ROA) is net profit divided by average total assets. It measures how efficiently the entire balance sheet is being put to work. A company with ₹500 of total assets earning ₹20 of profit has an ROA of 4 per cent.

One caveat before you start comparing numbers: stock screeners do not all calculate these the same way. Some use the year-end balance, others a two-year average; some define capital employed as total assets minus current liabilities, others as equity plus debt. The gaps are usually small, but when two sites show slightly different ratios for the same company, this is why.

The three numbers will rarely match for the same company, and the gap between them is itself a signal. A wide gap between ROE and ROCE means the company is using a lot of debt. A small ROA next to a flashy ROE almost always means leverage is doing the work, which is normal for banks and dangerous for everyone else.

How the three formulas actually work

Take Asian Paints as a worked example. In FY2024 (the year ended March 2024) the company earned about ₹5,558 crore of net profit on shareholders' equity of roughly ₹18,700 crore. That is an ROE of close to 30 per cent — thirty rupees of profit for every hundred rupees of owners' capital, a world-class number for a business that funds itself with almost no debt. (Screener)

Now look at capital employed. Because Asian Paints carries so little long-term debt, equity plus debt is barely more than equity alone, and its ROCE that year was actually higher than its ROE — around 40 per cent. When ROCE sits at or above ROE like this, you know the return is not borrowed: there is no leverage inflating the owners' slice.

The ROA tells the same story. Spread over a much larger asset base — which includes current liabilities like supplier payables — ROA still came in near 19 per cent. Across all three ratios the company looked like what it is: an efficient business using its capital well.

It is worth being honest about the latest year, because it is a lesson in itself. FY2025 was a weak one for Asian Paints: net profit fell roughly a third to about ₹3,710 crore, pulling ROE down into the high-teens-to-low-20s and ROCE to the mid-20s. Even excellent businesses have off years — which is exactly why you read five years of these ratios, not one.

Compare that with a hypothetical real-estate developer. Say it earns ₹100 crore on equity of ₹500 crore, giving an ROE of 20 per cent that looks good at first glance.

But the company carries ₹2,000 crore of debt against that equity. Capital employed is ₹2,500 crore, and operating profit before interest works out to maybe ₹250 crore. ROCE is only 10 per cent.

The story flips. The 20 per cent ROE was not a sign of a great business. It was a sign of a financially leveraged one, where a modest 10 per cent ROCE was magnified into a 20 per cent ROE by piling on debt. The day interest rates rise or sales slow, both ratios collapse together.

This leverage gap is the single most important pattern to learn. Pull up any Indian non-financial company on Screener. If ROE is significantly higher than ROCE, debt is doing the work, and the debt-to-equity ratio is your next stop. If they are close together, the business is genuinely productive.

For banks the math runs differently, and a picture explains why faster than a paragraph.

State Bank of India is the textbook case. In FY2025 it carried roughly ₹4.4 lakh crore of equity against about ₹66.8 lakh crore of total assets, so the owners' slice was under 7 per cent of the balance sheet. On a record net profit of ₹70,901 crore, that works out to an ROA of about 1.1 per cent — yet an ROE in the high teens (SBI reported 19.9 per cent on average equity). The ROE looks good only because the leverage is enormous, which is exactly how a bank is supposed to work. (SBI financial highlights)

When to use which ratio

Each ratio has a sector where it does its best work. Picking the wrong one is the same as judging a fish by how well it climbs trees.

For non-financial businesses, ROCE is the cleanest number to anchor on. It is the closest thing to an honest measure of operating efficiency. Sectors like FMCG, paints, IT services and quality manufacturing routinely throw up ROCE in the 25 to 40 per cent range. That is the band of genuine compounders.

For banks, NBFCs and insurance companies, the entire balance sheet is the product. ROA is the right anchor, because it shows the actual spread the institution earns on every rupee of asset. As an illustrative, through-the-cycle guide, strong private banks earn roughly 1.5 to 2 per cent ROA — the very best, like ICICI Bank and Kotak, have lately pushed past 2 per cent — while large PSU banks have climbed back to around 1 per cent in FY2025, after sitting far lower through the bad-loan years. These bands move with the credit cycle, so check the latest annual report rather than trusting a fixed range. (Screener)

ROE matters everywhere as the bottom-line shareholder return, but never read it alone. Always pair it with ROCE for non-financials and with ROA for financials.

FMCG names like Hindustan Unilever and Nestle India sit at the top of the table because they need almost no fixed assets to earn their profit. Asian Paints is in the same band. These businesses can fund their own growth from internal cash without taking on debt, which is exactly why their ROCE and ROE are both spectacular.

Capital-heavy sectors sit at the bottom not because they are bad businesses, but because they need a lot of money to operate. A cement company or a steel maker can be a perfectly fine investment at a 12 per cent ROCE; expecting it to match HUL is unrealistic.

The anchor question to ask is not "what is the absolute number" but "is this number good for this sector, and is it stable over five years."

The leverage trap that fools most investors

The single most common way ROE lies is through borrowed money. Replace one rupee of equity with one rupee of cheap debt and ROE goes up, even though the underlying business has not changed at all. The textbook calls this financial leverage. In practice it is the reason a lot of small-caps with 25 per cent ROE collapse the day the credit cycle turns.

Here is the test. Pull up any Indian non-financial stock on Screener and put ROE next to ROCE for the last five years. If the two numbers are close, the high return is real. If ROE is much higher than ROCE year after year, you are looking at leverage dressed up as quality.

Indian markets have given retail investors plenty of cautionary tales. DHFL was a fast-growing, profitable housing-finance lender for years — until its short-term wholesale borrowing caught up with it. After the IL&FS default froze the funding market in late 2018, DHFL defaulted in 2019, was sent into insolvency by the RBI, and was eventually bought out of bankruptcy by Piramal in 2021; the original shares were cancelled at zero and delisted. Leverage was not the only cause, but a balance sheet stretched that thin left no room for error. (Business Standard)

Many of the 2017-vintage small-cap infrastructure and real-estate names told a milder version of the same story: 20-plus per cent ROE during the easy-money phase, then halved or worse during the 2018 to 2020 deleveraging.

The opposite pattern is just as instructive. Tata Consultancy Services has run an ROE between roughly 30 and 50 per cent for over a decade with an almost debt-free balance sheet, reaching about 50 per cent in FY2024 and FY2025. Its ROCE is just as high — in fact higher — because there is no borrowing propping up the equity return at all. That kind of return is what compounding actually looks like. (Screener)

For banks, the same gap reverses meaning. A bank is supposed to be highly leveraged; that is its business model. So judging a bank on the ROE-versus-ROCE gap is meaningless. The right number to anchor on is ROA, which already accounts for the leverage built into the balance sheet.

HDFC Bank, after its 2023 merger with HDFC Ltd, earns an ROA of roughly 1.8 per cent and an ROE near 14 per cent — the marks of a genuinely well-run lender. A weaker bank limping along at a 0.5 per cent ROA, even with an optically reasonable 9 per cent ROE, is not the same animal. The ROA is the truth; the ROE is just leverage doing its job. (Equitymaster)

Once the leverage trap is visible, half the bad small-cap stories in the Indian market become easier to spot. Any non-financial company with a fat ROE-ROCE gap deserves a second look at the balance sheet before it deserves a place in the portfolio.

How to actually use the three ratios

Once the framework clicks, the workflow is simple and takes about ten minutes per stock.

Step one is the sector check. Decide whether the company is a financial or a non-financial. If it is a bank, NBFC or insurer, anchor on ROA. For everything else, anchor on ROCE.

Step two is the five-year trend. A single year of high ROCE or ROA can be a fluke. Five years of stable or rising numbers is a sign of a real franchise. Five years of declining numbers means the business is losing ground even if the latest snapshot still looks acceptable.

Step three is the peer comparison. Look at three or four other names in the same sector and see where the company sits. A 22 per cent ROCE looks great in isolation but is only average for a top-tier FMCG name. A 14 per cent ROCE is excellent for a cement company.

Step four is the leverage check. For non-financials, place ROE next to ROCE. A small gap means real returns. A large gap means debt is doing the work, and the next step is the debt-to-equity ratio.

Step five is the absolute floor. Anything earning less than its cost of capital — the minimum return investors expect, roughly 12 to 14 per cent for Indian companies, per the EY-NSE cost-of-capital survey — is destroying value over time, no matter how much absolute profit it shows. A non-financial business stuck permanently below that floor is a warning, not a value buy. (EY-NSE survey)

A practical exercise for a beginner is to spend an hour on a Saturday morning. Pick five companies you have heard of from completely different sectors. Note the current ROE, ROCE and ROA, the five-year average of each, and the debt-to-equity ratio.

The pattern jumps out fast. The genuine compounders, the leveraged frauds and the capital-heavy plodders sort themselves into three obvious groups. Doing this exercise on twenty companies is more useful than reading any number of brokerage reports.

Repeat it once a quarter as new results come in. Over a year, return ratios stop being numbers on a screener and become the way you actually understand which businesses are working and which are not.