Calculator

Owner Earnings Calculator: Buffett's True-Cash-Earnings Formula (Three Ways)

Net Income plus Depreciation minus Maintenance Capex — the exact formula Buffett published in 1986. The hard part is estimating maintenance capex. This tool runs three honest approximations side by side.

Vadim Kouznetsov·4 June 2026·7 min read

Why this owner earnings calculator exists

Owner earnings is Buffett's signature non-GAAP metric — the one number he says actually measures a business's true cash-generating power. He defined it precisely in his 1986 letter to Berkshire shareholders:

Owner earnings represent (a) reported earnings plus (b) depreciation, depletion, amortisation, and certain other non-cash charges, less (c) the average annual amount of capitalised expenditures for plant and equipment, etc., that the business requires to fully maintain its long-term competitive position and its unit volume.

The formula is simple. The problem is the last term. Maintenance capex — the cash a company must spend each year just to keep its existing business intact — is not a line item on any 10-K. Companies report total capex; analysts have to estimate how much of that maintains the existing business vs. how much funds growth (new factories, new stores, new technology).

Most online calculators dodge this entirely. They either ask you to enter maintenance capex manually (so the user has to know already), or they silently use total capex (which over-states the spending needed to maintain). Or they just use D&A as the only proxy (Buffett's original shortcut, but it doesn't work cleanly for every business).

This calculator does the honest thing: runs the formula three times with three different maintenance-capex assumptions and shows the range. If the three numbers cluster tightly, the business is mostly steady-state and you can trust any of them. If they're far apart, the company has meaningful growth capex or volatile capital needs, and you should think about which assumption fits your thesis.

How the math works

For each fiscal year of available history (typically 5-7 years from the 10-K filings), the calculator computes three versions of owner earnings:

Method 1 — D&A as maintenance capex (Buffett's traditional shortcut). When maintenance capex exactly equals depreciation, owner earnings collapse to GAAP net income. That's mathematically correct under Buffett's framework. The interpretation: in a stable business, depreciation is roughly the cash you need to spend each year to keep the existing asset base intact. If you book the same number on both sides, you arrive at net income as your owner-earnings approximation. This is Buffett's traditional default, and it works well for businesses where total capex ≈ D&A historically.

Method 2 — Total capex as maintenance capex (most conservative). This treats every dollar of capex as maintenance, implying the company has no growth capex. It's the most conservative interpretation and the closest analogue to free cash flow as it's typically reported. For mature, low-growth businesses this is often the right number. For growth businesses, it understates owner earnings (because some of that capex is investment in future earnings power, not maintenance of current earnings).

Method 3 — Custom percent of revenue (sanity check). The user enters a maintenance-capex assumption as a percentage of revenue. Default 5% (middle of the typical manufacturer range). Useful when methods 1 and 2 give very different answers — it lets you triangulate with a third reference point. Adjust to 1-3% for asset-light software businesses, 7-10% for capital-intensive industrials like utilities or telecom.

How to read the three-method spread

The three numbers told the same story would be a boring tool. They usually don't, and that's the value.

Narrow spread (all three methods within 10% of each other). The business is mostly steady-state. Total capex ≈ D&A, meaning the company is reinvesting roughly what it depreciates. Net income is a reasonable approximation of owner earnings. You can use any of the three numbers with confidence.

Total capex method is much lower than D&A method. The company has growth capex — total capex > D&A. The D&A method assumes only depreciation is "maintenance"; the rest of the capex is investing in future earnings power. Whether to trust this depends on whether the growth capex is generating real returns. For Microsoft and Costco today, growth capex is justified. For some industrial roll-ups, it's not.

Total capex method is much higher than D&A method. Rare — would mean total capex < D&A, which means the company is letting its asset base shrink. Either it's a deliberate wind-down or it's about to need a big catch-up capex cycle that will compress future owner earnings.

Custom % revenue method very different from both. Either the company's real maintenance capex is genuinely far from the simple D&A/total-capex framings, or your % assumption is off. Try adjusting the slider — at the right percentage, the method-3 number should land somewhere between the other two.

Implied P/E on owner earnings vs reported P/E

The most useful single output of this calculator is the implied P/E ratio computed against owner earnings, compared to the standard GAAP P/E quoted everywhere.

If a company has a reported P/E of 22× and an owner-earnings-based P/E of 19×, the business is cheaper than the headline suggests — its true cash earnings exceed reported earnings (because D&A overstates real maintenance capex). This is the case for asset-light software, payment networks, and brand-driven consumer staples.

If a company has a reported P/E of 18× and an owner-earnings-based P/E of 25×, the business is more expensive than the headline suggests — real cash earnings are lower than reported because growth capex is being capitalised on the balance sheet rather than expensed through the income statement. This is often the case for industrial roll-ups, capex-heavy real estate plays, and infrastructure builders.

Buffett's discipline says: pay attention to the owner-earnings P/E, not the GAAP P/E. The calculator surfaces both so you can compare directly.

Where this tool fits with the rest of the value-investing toolkit

Owner earnings is the cash number. To value the cash, you need a multiple or a discount rate. The companion tools handle that next step:

For the discounted-cash-flow valuation that turns owner earnings (or FCF) into an intrinsic value per share with a sensitivity table, see our free DCF calculator.
For the forward dividend math on whatever fraction of owner earnings the company returns as dividends, see our dividend calculator.
For the historical backtest of what those owner-earnings-driven returns actually compounded to in your portfolio, see our stock return calculator.

For the underlying value-investing discipline, the hub articles cover the framework: see owner earnings formula for the full derivation, margin of safety formula for how to apply Buffett's 25% discount to the intrinsic value the owner-earnings number drives, and how to know if a stock is a good buy for the broader checklist.

Limitations to be honest about

Owner earnings is only as good as the inputs. A few honest constraints:

Maintenance capex is genuinely uncertain. No method gives the "right" number. The three side-by-side approach is meant to surface that uncertainty, not paper over it. Treat any single number as an estimate.
Recent dips or spikes can distort the latest year. A one-off capex burst (data-centre buildout, acquisition spend booked as capex) inflates the conservative method's owner-earnings figure. The historical chart helps you see whether the latest year is anomalous.
Doesn't work for banks, insurers, or REITs. Their accounting reports cash flow differently. Use price-to-book and tangible book per share for banks; AFFO per share for REITs.
Newly-listed companies (under 3 years) may have insufficient history. The calculator requires at least one fiscal-year 10-K. Pre-IPO and recent-IPO names will return an error.
International ADRs work but the math is in USD. Currency translation effects in the underlying 20-F filings may cause some volatility year-over-year.