• 9 min. leestijd (EN)
ROE vs ROIC: Formula, Differences and When to Use
ROE and ROIC both answer the question "how well does this company use capital?" ROE looks at shareholders' equity alone, which makes it sensitive to leverage. ROIC looks at debt and equity together, which strips leverage out. The S&P 500 ROE heatmap shows ROE across the index at a glance; ROIC still requires pulling it from each company's filings or a separate data provider.
ROE vs ROIC at a Glance
The core difference: ROE rewards leverage, ROIC ignores it.
| Dimension | ROE | ROIC |
|---|---|---|
| Measures | Return on shareholders' equity only | Return on all invested capital (debt + equity) |
| Formula | Net Income / Shareholders' Equity | NOPAT / (Debt + Equity - Excess Cash) |
| Leverage | Boosted by debt and buybacks | Leverage-neutral |
| Typical range | 10-25% for S&P 500 companies | 6-30% depending on sector and capital intensity |
| Quality bar | Above 20% (Warren Buffett's classic screen) | 3 or more points above WACC (value creation) |
| Fails when | Equity approaches zero or turns negative | Operating and financing capital hard to split |
| Best for | Banks, insurers, equity-only analysis | Tech, industrial, retail, cross-sector compares |
What ROE Tells You
ROE divides net income by shareholders' equity. It answers: "For every dollar shareholders have left in the business, how many cents of profit did management generate this year?"
A company earning $500 million of net income on $2.5 billion of equity has an ROE of 20%. The S&P 500 long-term average sits near 14%, so 20% is the line many investors treat as exceptional. Warren Buffett's 1987 shareholder letter noted that only 25 of 1,000 large US companies averaged 20% ROE across 1977-1986 with no single year below 15%, and 24 of those 25 beat the S&P 500 over the same decade.
Limitation: ROE is blind to leverage. Borrow $1 billion, fund a buyback, and ROE climbs even if operations do not change. The DuPont breakdown exposes the split: ROE = net profit margin × asset turnover × equity multiplier. A rising equity multiplier means more debt, not better management. Always pair ROE with debt-to-equity before treating a high figure as quality.
Example: Apple's reported ROE has topped 150% in recent years, a number that mostly reflects years of aggressive share buybacks shrinking the equity base. Operating returns did not actually reach 150%.
What ROIC Tells You
ROIC divides Net Operating Profit After Tax (NOPAT) by invested capital, where invested capital equals debt plus equity minus excess cash and non-operating assets. NOPAT equals operating income (EBIT) multiplied by one minus the tax rate, so interest expense drops out of the numerator. Financing choices stop moving the ratio.
The metric answers: "For every dollar of productive capital inside the business, how many cents of operating profit does it earn after tax?" A company generating $2 billion NOPAT on $10 billion of invested capital has ROIC of 20%.
The useful signal is not the raw ROIC but the spread between ROIC and the weighted average cost of capital (WACC). If ROIC sits above WACC, each reinvested dollar creates value. If ROIC sits below WACC, growth destroys it. Morgan Stanley's Michael Mauboussin calls that spread the clearest single measure of a durable moat.
Limitation: ROIC is harder to compute than ROE. Data providers handle operating leases, goodwill, and cash balances differently, so published ROIC figures for the same company can vary by several percentage points.
Example: Costco's ROIC sits near 26%. Walmart's is closer to 16%. Similar scale, very different capital efficiency, and the gap partly explains why Costco trades at a premium multiple.
When to Use ROE vs ROIC
| Scenario | Better metric | Why |
|---|---|---|
| Banks and insurers | ROE | Leverage is structural; ROIC inputs are hard to isolate cleanly |
| Tech, industrial, consumer, retail | ROIC | Strips out capital-structure differences for fair peer comparison |
| Companies with heavy buybacks | ROIC | Shrinking equity inflates ROE without changing operations |
| Quick screen from headline figures | ROE | Net income and equity are on page one; NOPAT needs adjustments |
| Utilities, airlines, heavy industry | ROIC vs. WACC | Low ROIC is normal; only the spread above cost of capital matters |
| Cyclical companies | Both, 10-year | One-year numbers swing with the cycle on either metric |
| Startups and pre-profit firms | Neither | Both metrics break down without positive earnings |
Use ROE when: the business is a bank, insurer, or similar financial where leverage is part of operations, and a single equity-based figure captures the full picture.
Use ROIC when: comparing companies across industries or capital structures. A software business at ROIC 30% and a utility at ROIC 7% are not directly comparable on the absolute figure, but each is measurable against its own cost of capital.
Use neither alone when: earnings are negative, highly cyclical, or dominated by one-time items. Revenue-based alternatives like P/S or P/B give a steadier read in those cases.
Using ROE and ROIC Together
Run both numbers and watch the gap. If ROE is far higher than ROIC, leverage is doing the work. A company posting 25% ROE and 10% ROIC has borrowed its way to a headline figure, so pull the balance sheet and check debt-to-equity before concluding anything about operations. If ROE and ROIC sit close together, the returns are coming from operations, not financing.
Joel Greenblatt built his Magic Formula on ROIC for exactly this reason. The approach ranks stocks by a combination of ROIC (quality) and earnings yield (price), sorting for businesses that earn well and trade cheaply at the same time. Leverage tricks do not help a stock make that list.
A tighter screen for long-term compounders: require ROIC above 15%, ROE within 5 percentage points of ROIC, and both metrics stable across a full ten-year window that includes at least one recession. Those three conditions rule out debt-fuelled returns, single-year spikes, and cyclical peaks in one pass.
Pair the output with EPS growth across the same decade to confirm the high capital returns translate into earnings per share, not just flattering ratios.
Open the S&P 500 ROE heatmap to scan the index for names above the 20% bar, then pair each with a ROIC figure from the company's 10-K and a ten-year history before trusting any single number. ROIC is not on the heatmap yet, so that cross-check stays manual for now.
Frequently Asked Questions
Why do banks use ROE instead of ROIC? Banks are structurally leveraged: deposits, wholesale funding, and debt finance their lending. Separating operating capital from financing capital is close to impossible, so ROIC inputs become unreliable. ROE reflects the full business accurately for financials, which is why the industry has targeted it since the 1970s.
Can ROIC be higher than ROE? Yes, but rarely. It happens when a company carries large non-operating cash balances that the ROIC formula strips out, shrinking invested capital faster than it shrinks equity. If you see ROIC meaningfully above ROE, check the cash pile and the company's treatment of operating leases before drawing a conclusion.
Is ROIC the same as Return on Capital Employed (ROCE)? Close, not identical. ROIC uses NOPAT (operating profit after tax) divided by debt plus equity minus excess cash. ROCE uses EBIT divided by total capital employed without the tax adjustment and without stripping out cash. ROIC is cleaner for cross-company comparison. ROCE is easier to pull from a basic income statement.
What ROE and ROIC levels signal a quality business? Buffett's 20% ROE for ten years with no year below 15% is a high bar. A simpler ROIC rule: sustained ROIC above 15% with ROIC at least 3 percentage points above WACC marks a genuine moat. Anything below the cost of capital destroys value no matter how fast the company grows.