Owner earnings and free cash flow (FCF) are two cash-based attempts to answer the same question accrual accounting fudges: how much cash could an owner actually pull out of this business each year without weakening it? Both reject reported net income as the bottom line — net income carries non-cash charges and ignores the capital a business must spend to stay alive. The core tension between them is what counts as "necessary reinvestment": FCF subtracts all capital expenditure (a clean, reproducible number from the cash-flow statement), while Buffett's owner earnings subtracts only maintenance capex (the cash needed to hold competitive position), treating growth capex as discretionary. The first is objective but blunt; the second is more economically honest but requires an estimate that is, in Buffett's own words, "a guess."

How it's calculated

Free cash flow (the standard textbook form):

> FCF = Cash Flow from Operations − Capital Expenditures

This is "FCF to the firm" in its simplest retail form. Two more precise institutional variants exist (CFI, Wall Street Prep):

• FCFF (unlevered) = NOPAT + D&A − change in net working capital − Capex. Cash available to all capital providers, pre-debt.
• FCFE (levered) = FCFF + net borrowing − after-tax interest. Cash available to equity holders.

Owner earnings — Buffett, 1986 Berkshire Hathaway shareholder letter (verbatim structure):

> (a) reported earnings, plus (b) depreciation, depletion, amortization and certain other non-cash charges, less (c) the average annual amount of capitalized expenditures for plant and equipment, etc. that the business requires to fully maintain its long-term competitive position and its unit volume. If the business requires additional working capital to maintain its position, the increment also belongs in (c).

The load-bearing word is "maintain." Component (c) is maintenance capex, not total capex. The hardest input — separating maintenance from growth capex — has no GAAP line. The best-known estimation method is Bruce Greenwald's (in Value Investing): average the PP&E-to-sales ratio over ~5 years, multiply by the year's sales growth to isolate growth capex, then subtract growth capex from total capex to back into maintenance capex.

How it's used in practice

Both metrics serve three recognized jobs:

1. Valuation input. FCF (usually FCFF or FCFE) is the cash stream discounted in a discounted-cash-flow (DCF) model. Owner earnings is Buffett's preferred input to intrinsic-value estimation — he discounts owner earnings, not GAAP EPS. 2. Quality-of-earnings check. Comparing cash generation to reported net income flags accounting distortions. A company whose net income persistently exceeds its operating cash flow is converting paper profits, not cash — a classic red flag. 3. Capital-allocation capacity. FCF is the pool from which dividends, buybacks, debt paydown, and acquisitions are funded. "FCF yield" (FCF ÷ market cap or enterprise value) is a widely used valuation screen.

In practice analysts normalize: averaging capex over a full cycle, stripping one-off working-capital swings, and adjusting for items like capitalized software or operating leases. Owner earnings is rarely a single clean number — serious users present a range.

Standing & evidence

The deeper, evidence-backed idea underneath both metrics is that the cash component of earnings is more durable than the accrual component. This is documented, not folklore. Richard Sloan's 1996 paper ("the accruals anomaly") showed that firms with low or negative accruals (earnings backed by cash) subsequently outperformed high-accrual firms; a long-low/short-high accrual strategy earned roughly 10–12% annually in his sample, because the market overweighted accruals and underweighted their lower persistence (Sloan 1996; Stockopedia; Quantpedia). Importantly, the literature also reports the anomaly weakened after ~2002 as it became widely known and arbitraged (Green, Hand & Soliman; Quantpedia) — so the edge is real but decayed, and is a portfolio effect, not a single-stock guarantee. This supports cash-based metrics as a quality lens, but does not make a high-FCF stock a buy on its own.

Owner earnings itself has no measured standalone "edge" literature — it is a valuation philosophy, judged by the records of its practitioners, not a backtested signal. Treat efficacy claims about owner earnings as expert opinion, not proven anomaly.

Strengths & limitations

Strengths. Cash is far harder to fake than accrual earnings. Both metrics surface businesses whose reported profits aren't backed by cash, and both force the analyst to confront reinvestment needs that EPS hides.

The biggest live controversy — stock-based compensation (SBC). Because SBC is a non-cash charge, standard FCF adds it back, inflating FCF for stock-heavy companies. For large-cap SaaS firms, SBC has been estimated at roughly 39% of FCF versus ~4% for the S&P 500 (Wellington Management); a Morgan Stanley Counterpoint Global study of FCF-positive software firms put the median cash cost of SBC near 99% of FCF — so the distortion is largest exactly where these stocks trade. Buffett's view is blunt: "The very name says it all: 'compensation.' If compensation isn't an expense, what is it?" (1998 Berkshire letter). SBC is a real, recurring cost that dilutes owners; treating add-back FCF as "owner cash" overstates returns unless you also subtract the buybacks used to offset dilution. This is the single most common modern misuse of FCF.

Other failure modes:

• Maintenance-capex guesswork. Owner earnings' headline strength is also its weakness — (c) is an estimate. Two honest analysts can produce materially different owner-earnings figures for the same firm.
• Lumpy capex. A single big plant or one-time growth year makes one-year FCF meaningless; you must normalize over a cycle.
• Working-capital games. Stretching payables or under-investing in inventory boosts cash flow temporarily while quietly damaging the business — "good" FCF that's actually a warning.
• Negative ≠ bad. A healthy company reinvesting heavily for growth can run negative FCF for years (early Amazon); the metric punishes growth investment it can't distinguish from waste.

Sources

• Warren Buffett, 1986 Berkshire Hathaway Shareholder Letter (owner-earnings definition, appendix) — primary source.
• Wikipedia, "Owner earnings" — formula, Greenwald maintenance-capex method, Buffett's "must be a guess" caveat.
• Corporate Finance Institute; Wall Street Prep — FCF, FCFF, FCFE formulas.
• Richard Sloan (1996), "Do Stock Prices Fully Reflect Information in Accruals and Cash Flows About Future Earnings?" via Stockopedia and Quantpedia — accruals anomaly (~10–12% historical spread; post-2002 decay).
• Wellington Management, "Stock-based compensation for tech sector" — SBC ≈ 39% of FCF for large-cap SaaS vs ~4% for the S&P 500.
• Morgan Stanley Counterpoint Global Insights, "Stock-Based Compensation: Unpacking the Issues" — the add-back distortion; median SBC ≈ 99% of FCF across FCF-positive software firms.
• Warren Buffett, 1998 Berkshire Hathaway Shareholder Letter — "If compensation isn't an expense, what is it?"
• Old School Value; StableBread — owner-earnings worked-calculation walkthroughs (cross-check).

Disputes flagged: (1) SBC treatment — Buffett vs. the "SBC is a non-cash financing item" camp — is genuinely contested. (2) Owner earnings has no measured standalone edge; only the underlying cash-vs-accrual persistence is academically supported, and that has weakened post-2002.