Analyzing StocksStep 2.12 · article 18 of 32 in the learning path
Intermediate · 55/1008 min readTopic: Profitability
ROIC vs ROE vs ROI: The Complete Guide to the Three Return Metrics
ROI, ROE and ROIC all measure return on money, but each uses a different definition of “money in”. This guide explains the difference with worked examples — in beginner, intermediate and advanced versions — and shows why ROIC, not ROE, is the metric most tied to long-term value creation.
New to these? Tap any term for a plain-English definition and example.
Showing your level:
ROI, ROE and ROIC are three ways of answering one simple question: for every dollar put in, how many cents came back? They look similar, but each measures a different "dollar in".
The one-sentence version
ROI — Return on Investment. The all-purpose "did this pay off?" number for any decision.
ROE — Return on Equity. How well a company turns shareholders' money into profit.
ROIC — Return on Invested Capital. How well a company turns all the money in the business (from shareholders and lenders) into profit.
A lemonade-stand example
Imagine you open a lemonade stand.
You spend $100 on a stand, cups and lemons.
Over the summer you earn $120 back.
Your profit is $20.
Your ROI is profit ÷ money in = $20 ÷ $100 = 20%. Simple.
Now say you only had $60 of your own money, so you borrowed $40 from a friend. The $60 you personally put in is your equity. If the stand still made $20 profit, your return on your own money (a rough ROE idea) looks bigger, because you used a smaller slice of your own cash. The other $40 came from someone else.
ROIC looks at the whole $100 working in the business — your $60 plus the friend's $40 — and asks how good the lemonade stand itself is, ignoring who paid for it.
The key idea: ROE can look great just because a company borrowed a lot. ROIC ignores borrowing and shows how good the underlying business really is.
Quick comparison
Metric
"Money in" is…
Best for
ROI
Whatever you spent on one thing
Any project, ad campaign, or purchase
ROE
Shareholders' money only
Returns to the owners of a company
ROIC
Shareholders' + lenders' money
How good the core business is
What "good" looks like
As a rough rule of thumb for a healthy company, an ROIC or ROE in the mid-teens or higher (say 15%+) sustained for years is a sign of a strong, efficient business. A single year tells you little — look for a steady track record.
Takeaways
All three say "cents back per dollar in" — they only differ on which dollars.
Use ROI for one-off decisions, ROE for owner returns, ROIC to judge the business itself.
If ROE looks high but ROIC is low, the company is probably leaning on debt.
All three ratios divide a profit figure by a capital figure. The art is matching the right profit to the right capital — mixing them up is the most common analysis mistake.
The formulas
Metric
Formula
Captures financing effects?
ROI
(Gain − Cost) ÷ Cost
N/A — general purpose
ROE
Net Income ÷ Shareholders' Equity
Yes — leverage flatters it
ROIC
NOPAT ÷ Invested Capital
No — capital-structure neutral
Where NOPAT = Net Operating Profit After Tax = Operating Income × (1 − tax rate), and Invested Capital ≈ total debt + equity − excess cash.
Why ROE and ROIC can disagree: a worked example
Two companies each run an identical business that produces $100 of operating profit on $1,000 of invested capital. Assume a 20% tax rate, so NOPAT = $80 and ROIC = 80 ÷ 1,000 = 8% for both.
Company A (no debt)
Company B (50% debt at 5%)
Invested capital
$1,000
$1,000
Debt / Equity
$0 / $1,000
$500 / $500
Operating profit
$100
$100
Interest (5% × debt)
$0
−$25
Pre-tax profit
$100
$75
Net income (after 20% tax)
$80
$60
ROIC
8%
8%
ROE
80 / 1,000 = 8%
60 / 500 = 12%
Same business, same operating efficiency (ROIC 8%), yet Company B reports ROE of 12%. The extra 4 points are pure leverage, not better operations — and that leverage also adds risk.
DuPont: where does ROE come from?
The DuPont identity splits ROE into three drivers, so you can see whether a high ROE is "earned" or "borrowed":
ROE = Net Margin × Asset Turnover × Equity Multiplier
The third term, the equity multiplier, rises with debt. If two firms have similar margins and turnover but very different ROEs, the gap is leverage.
The number that decides value: ROIC vs WACC
A company creates value only when ROIC > WACC (its weighted average cost of capital). If ROIC < WACC, growth actually destroys value — every extra dollar invested returns less than it costs to fund. This is why ROIC, not ROE, is the metric most tied to long-term value creation.
When to use which
ROI — evaluating a single project, campaign, acquisition, or capital expenditure.
ROE — gauging returns to shareholders; standard for banks and insurers where leverage is the business model.
ROIC — comparing the underlying quality of two operating businesses with different debt levels; pairing against WACC.
Common pitfalls
Leverage mirage: a soaring ROE driven by buybacks or debt, while ROIC is flat.
Negative equity: heavy buybacks can push equity near or below zero, making ROE meaningless.
One-off profits: asset sales or tax windfalls inflate the numerator for a year.
Cash distortion: a large idle cash pile drags ROIC down unless you exclude excess cash from invested capital.
At an analyst level the three ratios converge on one discipline: build a numerator and denominator that are consistent (operating-with-operating, after-tax-with-after-tax) and normalized for distortions, then judge the result against the cost of capital.
Defining the components precisely
NOPAT
NOPAT isolates after-tax operating profit, independent of financing:
NOPAT = EBIT × (1 − cash tax rate)
Refinements analysts make: add back the implied interest on capitalized operating leases (post-IFRS 16 / ASC 842 most leases are already on balance sheet), capitalize and amortize R&D where it is truly investment, and use a normalized/cash tax rate rather than the headline rate.
Operating approach: net working capital + net PP&E + capitalized intangibles + goodwill.
Decide deliberately whether to include goodwill: with goodwill measures returns on the price actually paid for acquisitions (a discipline check on M&A); without goodwill measures the operating return of the underlying assets. Report both for serial acquirers.
DuPont, extended
The five-step DuPont separates operating performance from financing and tax:
This lets you attribute an ROE change to operations (margin × turnover), the cost/level of debt (interest burden × equity multiplier) or tax. A useful bridge:
So ROE = ROIC plus a leverage spread. Positive spread × more debt = higher ROE and higher risk; if ROIC falls below the after-tax cost of debt, leverage amplifies losses.
Economic profit: the ROIC–WACC spread
Value creation is the spread, scaled by the capital deployed:
Economic Profit (EVA) = (ROIC − WACC) × Invested Capital
A firm can grow earnings every year and still destroy value if ROIC < WACC. Conversely a slow-growing firm with a wide, durable ROIC–WACC spread compounds intrinsic value. This spread — its size and, above all, its durability (the "moat") — is what links ROIC to valuation.
Growth, reinvestment and incremental ROIC
Sustainable growth is funded by reinvesting at the marginal return:
g = Reinvestment Rate × ROIC
Headline (average) ROIC reflects the legacy asset base; incremental ROIC — the return on newly invested capital, ΔNOPAT ÷ ΔInvested Capital — reveals whether fresh investment still earns above the cost of capital. A high average ROIC masking falling incremental ROIC is an early warning that growth is getting expensive.
Normalizations and traps
Distortion
Effect
Fix
Buybacks / negative equity
ROE explodes or turns meaningless
Prefer ROIC; inspect equity trend
Large excess cash
Understated ROIC
Exclude excess cash from invested capital
Acquisition goodwill
Two very different ROICs
Report ROIC with and without goodwill
Operating leases
Off-balance-sheet capital (legacy GAAP)
Capitalize leases; adjust NOPAT for implied interest
R&D / brand spend
Investment expensed, capital understated
Capitalize & amortize where appropriate
One-off items
Numerator spikes for a year
Use normalized, multi-year figures
Sector nuance
Banks & insurers: leverage is the model and "invested capital" is ill-defined — use ROE and ROA against regulatory capital, not ROIC.
Capital-light (software, brands): invested capital is small, so ROIC can be enormous or even negative-denominator; focus on incremental ROIC and cash conversion.
Capital-heavy (utilities, telecom, industrials): ROIC vs WACC is the central lens; small spread changes move valuation materially.
Analyst's checklist: consistent numerator/denominator → normalize one-offs, leases, R&D, excess cash → show ROIC with/without goodwill → compare ROIC to WACC → check the incremental ROIC and the durability of the spread.
Test what you've learned
7 quiz questions cover this topic, each with an instant explanation.
We use cookies to run the site and, only if you allow it, for live chat, analytics, personalization and ad measurement.
We use cookies to:
Deliver and maintain our services, protect against spam, fraud and abuse
Offer a live chat with our team (Enuchat, optional)
Measure audience engagement and site statistics to understand how our services are used (Google Analytics, optional)
Remember your quiz progress and tailor the knowledge base to your skill level (optional)
Measure how our ads on Facebook and Instagram perform, including whether a visit leads to a confirmed sign-up, matched to your Facebook or Instagram account using hashed contact details you enter (Meta Pixel, optional)
Optional cookies are only set if you allow them. You can accept all, reject all or choose per category, and change your choice at any time via the "Manage Cookies" link in the footer. Google and Meta may process this data in the USA. More in our Privacy Policy.
Manage cookie preferences
Necessary cookiesAlways active
Required for the website to function. These cannot be disabled.
Cookie
Purpose
Duration
PHPSESSID
Session identifier for login, security tokens, and site functionality
Session
cookie_consent
Stores your cookie preferences
1 year
Functional cookies
Show the live chat button so you can message our team. The Enuchat chat widget stores a visitor ID in your browser and sends your chat messages to Enuchat. When this is off, the chat is not loaded.
Cookie
Purpose
Duration
enuchat_visitor_id
Keeps your chat conversation linked to this browser (Enuchat, browser storage)
Until you clear it or turn this off
Analytics cookies
Help us understand how visitors use our website with Google Analytics, provided by Google LLC. Google receives pages visited, device data and an IP address, and may process them in the USA. If you also allow marketing cookies, the Google tag can detect contact details you enter or see on our pages, such as your email address, phone number, name or address, hash them before sending and pass them to Google so it can match your visits to a Google account for measuring our ads. Hashed details are still personal data. See how Google uses information from sites that use its services.
Cookie
Purpose
Duration
_ga
Distinguishes unique visitors (Google Analytics)
2 years
_ga_*
Maintains session state (Google Analytics)
2 years
Personalization cookies
Remember your quiz progress and work out your investing level so we can recommend knowledge-base articles and the right reading depth for you. Stored only on your device.
Cookie
Purpose
Duration
bes_user_level
Your quiz-derived investing level, used to tailor content
1 year
Marketing cookies
Let Meta (Facebook and Instagram) measure how our ads perform, using the Meta Pixel. Meta Platforms Ireland Ltd. receives the pages you visit and, when you confirm the email address of a new account, a "completed registration" event, together with browser identifiers and your IP address. Meta's automatic advanced matching also reads contact details you type into forms on our site, such as your email address or name, hashes them in your browser and sends them with these events so Meta can match them to a Facebook or Instagram account. Hashed details are still personal data. These events are sent from your browser and may also be forwarded to Meta through Meta's Conversions API Gateway. Our own server also sends page visits and completed-registration events directly to Meta's Conversions API, with your IP address, browser user agent and Meta browser identifiers, so Meta can count it once. To help Meta match these visits, your browser and our server also send a random visitor ID from our own cookie and, if you are signed in, your account ID, your confirmed email address and the name, city, postal code and country saved in your profile, all hashed first. We and Meta are joint controllers for collecting and sending this data, and Meta's further use, which may include processing in the USA, is described in the Meta Privacy Policy. When this is off, the Meta Pixel is not loaded and no page visit or registration event is sent to Meta. Together with analytics cookies, this choice also lets Google receive hashed contact details, as described under Analytics cookies.
Cookie
Purpose
Duration
_fbp
Identifies browsers for Meta ad measurement (Meta Pixel)
3 months
_fbc
Stores the Meta ad click that brought you here (Meta Pixel)
3 months
_bes_meid
Random visitor ID, sent to Meta only hashed, so your visits can be matched (Meta Pixel and Conversions API)