The Value Toolkit · Module 09

Owner Earnings

Warren Buffett doesn't trust reported earnings — and he told the world why in 1986. The profit a company prints is an accounting number, full of non-cash charges and silent about the cash it must spend just to stand still. Owner earnings strips all that away to ask the only question an owner cares about: how much cash could I actually take out each year without harming the business?

~7 min readFor the experienced investorQuick reference: Methodology →

01Where it comes from

In the appendix to his 1986 Berkshire Hathaway shareholder letter, Buffett took aim at a sacred number: reported net income. Depreciation, he argued, is a real economic cost for some businesses and a near-fiction for others; meanwhile the capital a company must reinvest just to hold its position never even appears on the income statement. So the earnings line can flatter a business that's quietly bleeding cash into its own machinery.

His fix was a concept he called owner earnings— the cash a rational owner could pull out of the business each year while leaving its competitive position intact. It's Buffett's honest answer to “what does this company really earn?”, and it's the spirit behind every discounted-cash-flow model that came after.

Buffett was picking a specific fight. Through the 1980s, dealmakers had fallen in love with “cash flow” and EBITDA — earnings before all the inconvenient parts — to justify ever-higher takeover prices. His retort was blunt: any measure that ignores capital spending is worse than useless, because for most businesses that reinvestment is as real and unavoidable a cost as payroll. Companies that skimp on it don't stay the same size — they quietly shrink. Owner earnings was his insistence that you subtract the cash a business must genuinely plow back just to hold its ground, and count only what's truly left for the owner. The maddening part, which he freely admitted, is that this number is the one figure companies never actually report — so estimating it is an act of judgment, not arithmetic.

02The formula

Owner Earnings
Net income + Depreciation & amortization Maintenance capex ± Change in working capital

Add back D&Abecause it's a non-cash charge — the cash left the business years ago when the asset was bought. Subtract maintenance capex — the real, ongoing cash needed to keep the business running as-is. Adjust for working capitalbecause cash tied up in inventory and receivables isn't cash you can pocket. What remains is the true distributable cash.

The whole exercise is a bridge from an accountant's number to an owner's number. Here it is on an illustrative company, per share, so the moving parts are clear:

Reported earnings → owner earnings
Illustrative, per share — round numbers to show the bridge, not a real company.
Net income (reported EPS)$5.00
+ Depreciation & amortization (non-cash, added back)+ $3.00
− Maintenance capex (cash to stand still)− $2.00
− Increase in working capital− $0.50
= Owner earnings$5.50

Here owner earnings ($5.50) landed a little above reported EPS ($5.00) — depreciation exceeded the cash actually needed to reinvest. For a capital-light software firm the gap can be enormous and positive; for an airline or a steelmaker, owner earnings can sit far below reported profit. That gap is the whole point. The app applies a rough ~10× owner-earnings ceiling as a fair-value gut check.

03On a real business — and why it's a range

Linamar, our through-line stock, shows exactly where owner earnings gets slippery:

LinamarTSX: LNR.TO
Owner-earnings inputs & price: Aug 23, 2026 snapshot (last close) — not live data
Reported EPS≈ $11.60
Owner earnings / sh≈ $13 – $24
Price ÷ owner earnings≈ 4.5× – 8.4×

Because Linamar's depreciation is heavy, owner earnings run above its reported $11.60 EPS. But the exact figure depends on one hard judgment: how much of its capital spending is maintenance (keeping the plants running) versus growth(building new capacity). Count all capex as maintenance and owner earnings are ~$13/share; strip out growth capex and it's ~$24. At ~$109, that's between 4.5× and 8.4× owner earnings — comfortably under the ~10× ceiling either way. The app deliberately uses total capex (the conservative end), so it never flatters a company by guessing its growth spending.

04When owner earnings lie

⚠ Where the judgment hides

Three soft spots

  • Maintenance capex is a guess.Companies don't report the split between “keep the lights on” and “expand,” so every owner-earnings figure embeds an assumption — and small changes to it swing the answer a lot (the $13-vs-$24 gap above). Treat it as a range, not a point.
  • One year of working capital can distort it. A single swing in inventory or receivables — a stock-up before a big year, a one-time collection — can move owner earnings far more than the underlying business changed. Look across several years.
  • It's not for financials.Banks and insurers don't have “capex” or working capital in the ordinary sense, so the bridge doesn't apply — their cash generation runs on entirely different rails.

Owner earnings is less a precise number than a discipline— a refusal to take reported profit at face value. That's why the app shows it as a rough ceiling beside the harder-edged methods rather than as a single “fair value,” and leans to the conservative side of the capex question. Used that way, it's the closest thing to seeing a business the way its owner does.

◆ Run it yourself

See owner earnings on any stock

The research page computes owner earnings from the latest filings and shows the ~10× ceiling beside the Graham methods, EPV, and the DCF — so you can see the accountant's number and the owner's number side by side, recomputed from current data.

Source: Warren Buffett, 1986 Berkshire Hathaway shareholder letter (appendix), which introduced “owner earnings.” Owner-earnings inputs and price for LNR.TO are from the Aug 23, 2026 snapshot (last market close) and are not live; the illustrative bridge uses invented round numbers. The live research page recomputes owner earnings from the most recent filings, using total capex (the conservative choice).

For educational use. This is not financial, investment, or tax advice, and nothing here is a recommendation to buy or sell any security. Every model has known blind spots — always verify against a company's primary filings before acting. · travisvaluation.ca