Hold on to this
A 20% ROIC sounds impressive. A 20% ROIC paired with a 22% WACC is a company destroying value on every dollar it reinvests.
Instrument · Drive it
Try it yourself
ROIC vs. WACC Spread: The Value-Creation Test
The spread — ROIC minus WACC — is economic profit earned on every dollar of capital deployed: a positive spread compounds shareholder wealth, a negative one erodes it. When the rate rises of 2022–2023 pushed WACC up 200–300 basis points, capital-intensive sectors like utilities, industrials, and real estate saw their spreads compress or turn negative even with stable ROIC, because their heavy asset bases carry a higher cost of capital. Asset-light businesses — software, consumer internet — held positive spreads through the same shock, since low capital intensity insulates their returns from the cost of money.
| Industry | Typical ROIC | Typical WACC | Spread | Verdict |
|---|---|---|---|---|
| Software / SaaS | 22% | 10% | +12% | Value Creating |
| Semiconductors | 18% | 11% | +7% | Value Creating |
| Medical Devices | 16% | 9% | +7% | Value Creating |
| Consumer Staples | 15% | 8% | +7% | Value Creating |
| Specialty Retail | 13% | 9% | +4% | Value Creating |
| Pharmaceuticals | 13% | 9% | +4% | Value Creating |
| Aerospace & Defense | 12% | 9% | +3% | Value Creating |
| Healthcare Services | 11% | 9% | +2% | Value Creating |
| Asset Management | 11% | 9% | +2% | Value Creating |
| Industrials | 10% | 9% | +1% | Value Creating |
| Auto & Components | 9% | 10% | −1% | Value Destroying |
| E-Commerce / Internet | 9% | 10% | −1% | Value Destroying |
| Oil & Gas Exploration | 9% | 11% | −2% | Value Destroying |
| Chemicals | 8% | 9% | −1% | Value Destroying |
| Telecom | 8% | 9% | −1% | Value Destroying |
| Metals & Mining | 8% | 10% | −2% | Value Destroying |
| Real Estate (REITs) | 7% | 9% | −2% | Value Destroying |
| Airlines | 7% | 10% | −3% | Value Destroying |
| Electric Utilities | 7% | 8% | −1% | Value Destroying |
| Steel & Materials | 6% | 10% | −4% | Value Destroying |
Typical mid-cycle estimates; WACC reflects post-2022 rate environment. Spread = Typical ROIC − Typical WACC.
Move from benchmarks to your own numbers: calculate your exact WACC, compute ROIC for any ticker, then run a DCF using your spread estimate.
What ROIC actually measures
Strip away the formula and ROIC asks one question: for every dollar management put to work inside this business last year, how many cents came back as operating profit? That is harder to answer — and harder to game — than earnings per share or return on equity.
ROE leaves out debt. A company that borrows $2 billion to buy back shares shrinks the equity base, and ROE climbs without the business earning a single extra dollar. ROIC holds steady because the new debt flows straight into invested capital — the denominator expands to match the financing. When ROE and ROIC diverge sharply, one ratio is measuring capital efficiency and the other is measuring capital structure.
ROA divides by total assets, which includes liabilities suppliers fund for free. A company extracting 90-day payment terms from its vendors looks capital-heavy on ROA but is actually quite efficient — it finances part of its operation at zero cost. Invested capital strips those out, leaving only the money that had a real cost attached to it.
ROIC consistently above the cost of capital is the arithmetic definition of a competitive moat. Companies that sustain ROIC above their hurdle rate for five-plus years tend to create more shareholder wealth than those with faster EPS growth but lower capital efficiency — because each retained dollar is being put to better use on the margin.
The formula — and where people get it wrong
ROIC = NOPAT ÷ Invested Capital. Neither term appears as a labeled line on a standard 10-K. You build both from pieces that do.
NOPAT starts with operating income (EBIT) — the income statement usually labels this directly — and multiplies by one minus the effective tax rate. The result is the operating profit available to all capital providers, stripped of financing costs. Using EBIT rather than net income keeps the capital structure out of the numerator, so the ratio is comparable across companies with different leverage.
Invested capital takes total assets, subtracts non-interest-bearing current liabilities (accounts payable, accrued expenses, deferred revenue — anything the company does not pay interest on), and subtracts excess cash. The reconciliation is the reassuring part: you can reach the same figure from the financing side — equity plus interest-bearing debt minus excess cash — and if the two approaches disagree, you missed a line item.
Three decisions trip most analysts. First, goodwill: include it. Goodwill is real cash paid above book value for an acquisition. Keeping it in the denominator tests whether management priced the deal well — stripping it inflates ROIC and conceals capital-allocation discipline. Some analysts compute both: goodwill-inclusive tests total return on deployed capital, goodwill-exclusive tests organic earning power. Either is valid, but you have to be explicit before comparing across companies.
Second, operating leases. Under ASC 842, the right-of-use asset already sits on the balance sheet for post-2019 US GAAP filings — it is already in invested capital. For older data or peers on different standards, capitalize manually by multiplying the annual lease expense by roughly 6–8× as a proxy for present value.
Third, excess cash. A company operating normally keeps roughly 1–2% of revenue in cash. Anything above that is a parking decision, not an operating one. Leaving $3 billion of idle cash in the denominator of a company whose core business earns 30% ROIC dilutes the signal on the business you are actually trying to evaluate.
The spread: ROIC minus WACC
ROIC in isolation tells you nothing. A 20% ROIC paired with a 22% WACC is a company destroying value on every dollar it reinvests. A 10% ROIC paired with a 7% WACC is compounding economic wealth at 3 points per year. The spread — ROIC minus WACC — is the operative signal, and the analyzer above forces you to input both numbers to see it.
WACC blends the after-tax cost of debt with the expected return on equity, weighted by capital structure. For most public companies it falls between 7% and 12%. A regulated utility in a low-rate environment might carry a 6% WACC; a small-cap tech company with no debt and a high beta might face 13%. The hurdle rate is different for every company. Comparing raw ROIC across industries without anchoring to cost of capital is not just imprecise — it can be backwards in direction.
The compounding effect of a sustained positive spread is what drives long-run outperformance. See's Candies, acquired by Berkshire Hathaway in 1972, earned roughly 25% ROIC against a 10–12% cost of capital for decades. That 13–15 point spread, held for years, produced compounding far beyond what the business's modest size implied. Pair the spread calculation with the WACC guide to estimate the hurdle rate from scratch for any company you are evaluating.
What counts as good — by industry
There is no universal threshold. The right benchmark is whether ROIC clears the company's own WACC consistently — not just in a cyclical peak year. Within that, industry structure sets the range.
Capital-light businesses with pricing power — software platforms, payment networks, branded consumer goods — routinely post 20–60% ROIC. Their invested capital base is small relative to earnings, and incremental revenue carries high margins. Microsoft exceeded 40% in recent fiscal years. Visa runs above 30% consistently. These are not outliers; they reflect the economics of businesses that grow revenue without proportionally adding to the asset base.
Capital-intensive businesses operate under different math. Regulated utilities earn 7–9% by design — regulators set allowed returns and heavy infrastructure spending keeps invested capital growing. Airlines have historically earned below their cost of capital. That is the arithmetic explanation for why the industry has collectively destroyed more wealth than it has created since deregulation.
- Software (SaaS): 20–60%. Invested capital is thin, margins are high, retention compounds the base. High ROIC here is structural, not cyclical.
- Specialty distribution: 15–30% for the best operators. AutoZone (AZO) and Fastenal reach the high end via inventory discipline and pricing power.
- Branded healthcare and pharma: 10–25%. Wide range — branded products compound well; generic manufacturers compete on thin spreads.
- Industrials and materials: 8–15% in normal conditions, highly cyclical. A 14% ROIC at a commodity peak may represent 6% through the cycle.
- Utilities and regulated telecom: 5–9%. Value creation depends on the spread above WACC, not the absolute ROIC level — the regulatory framework almost always constrains the ceiling.
Trend over time beats any single year
One year of ROIC is a snapshot. Five years is a diagnosis. A business that held 17–20% ROIC across different macro environments has demonstrated structural earning power. One that posted 22%, then 17%, then 13% is showing compression — and the cause matters more than the number.
M&A dilution is the most common culprit. An acquisition at a significant premium loads goodwill onto invested capital immediately, while the acquired earnings take time to flow through NOPAT. ROIC will compress for 12–24 months after any large deal — that is normal. The signal is the recovery: if ROIC does not trend back toward pre-deal levels within two to three years, the acquirer overpaid or failed to extract the synergies that justified the premium.
The second pattern is growth into declining returns. A company expanding its capital base at 15% per year while NOPAT grows at 8% is systematically deploying capital into worse projects. This often precedes a restructuring by several years — the economic damage is happening in real time while the accounting impact is deferred. NVR Inc. (NVR) is the counterexample: by using land options rather than owning land outright, the homebuilder keeps invested capital permanently thin and ROIC consistently in the 30–40% range across full housing cycles. A deliberate capital-structure choice, not a lucky outcome.
Compare five-year ROIC trends against sector peers on the same time window. A company that compressed 4 points while the sector compressed 8 may be gaining competitive position. One that compressed 4 points while peers gained 2 has a company-specific problem. The delta against sector context tells you more than the absolute level does.
Distortions that inflate or deflate the reported number
Several accounting events move ROIC in ways that mislead if read without context.
- Goodwill impairment: When goodwill is written down, invested capital shrinks immediately. NOPAT is also reduced in the impairment year but recovers the next year while the smaller denominator persists. ROIC can jump mechanically after a write-down even though the underlying business is unchanged. The check: did NOPAT grow, or did the base just get smaller?
- R&D expensing: GAAP expenses R&D immediately, keeping invested capital artificially thin — those assets never accumulate on the balance sheet. Technology companies can print 80–100%+ ROIC partly because decades of expensed R&D never appear in the denominator. This is partly real (the model is genuinely capital-light) and partly an accounting artifact.
- Fresh acquisitions: A large deal immediately loads goodwill into invested capital, compressing ROIC even if the deal was priced fairly. The relevant question is the recovery trajectory over the following two to three years, not the quarter the deal closed.
- One-time charges in EBIT: Restructuring charges, litigation settlements, and impairments flow through operating income. A company can look worse in the year it cleans house than in the year it was actually deteriorating. Adjust for genuine one-time items — but do it consistently across all years. A company that restructures every three years is not incurring a non-recurring charge.
Cross-check with free cash flow trends to confirm that operating profit is converting to cash at a normal rate, and with earnings quality metrics to verify NOPAT is not inflated by aggressive accruals. A high ROIC backed by strong FCF conversion is a materially stronger signal than one that is not.
Questions worth asking
What is a good ROIC?
Above 15% consistently suggests durable competitive advantage, but the more useful test is whether ROIC exceeds WACC. A 10% ROIC in a business with an 8% cost of capital creates value. A 15% ROIC with a 16% cost of capital does not. Industry context matters too — software companies routinely hit 30%+, while capital-intensive industries may be doing well at 8%.
How is ROIC different from ROE?
ROE only measures returns on equity, so a company that loads up on debt can inflate ROE without becoming a better business. ROIC uses total invested capital — debt plus equity, minus non-operating cash — which makes it far harder to game and much more comparable across companies with different capital structures.
Should goodwill be included in invested capital?
Usually yes. Goodwill represents real cash paid above book value for an acquisition — including it tests whether management overpaid. Some analysts strip it out to assess organic earning power, which is a valid exercise. Calculate both and be clear which question you're answering before comparing across companies.
Why do software companies have such high ROIC?
Their invested capital base is small relative to earnings. Software has minimal physical assets, R&D is expensed (which lowers NOPAT but also shrinks invested capital), and incremental revenue carries very high margins. A SaaS company at 30% ROIC is not doing something extraordinary — it reflects the economics of the model.
What does declining ROIC usually signal?
Often one of three things: margins compressing (pricing power loss), acquisitions at high prices (goodwill inflating the capital base), or growth capital being deployed into lower-return projects. Declining ROIC alongside accelerating capital deployment is the most bearish combination — the company is growing into worse returns.
Keep reading