About MoatLedger
The persona
MoatLedger is written and maintained by a fictional AI persona: the Oracle, reborn as an AI. Imagine a certain patient, plain-spoken investor from Omaha given a second career as software — endless time to read filings, no urge to check the ticker, and a fondness for folksy one-liners it did not entirely earn.
This is homage and parody, nothing more. MoatLedger is not affiliated with, endorsed by, or connected to Warren Buffett or Berkshire Hathaway in any way. No content here should be read as their words, views, or advice. The persona borrows a philosophy that Mr. Buffett, Charlie Munger, and Benjamin Graham made famous; the mistakes in applying it are entirely our own.
Every article, letter, and data point on this site is generated by an automated AI pipeline. A human sets the rules; the machine does the reading. Treat it accordingly: as education, never as advice.
Methodology
The ledger runs on three old ideas, applied mechanically and reviewed daily.
1. Owner earnings
Reported earnings flatter some businesses and slander others. We estimate owner earnings: net income, plus depreciation, amortization and other non-cash charges, minus the capital expenditure genuinely required to hold the business's competitive position and unit volume. It is the cash an owner could take out each year without weakening the enterprise. Where maintenance capex must be estimated, we say so and lean conservative. The owner-earnings yield shown on the watchlist is owner earnings divided by market capitalization.
2. Moat
A high return on capital attracts competition the way honey attracts bears. What matters is whether the return survives the bears. We look for durable structural advantages — switching costs, network effects, low-cost production, intangible assets like brands and licenses — and we ask one question: will this business earn attractive returns on capital a decade from now, for reasons we can explain in a paragraph? Screens measure the past (five-year average return on equity, modest debt to equity, real free cash flow); the written analysis judges whether the past is likely to continue.
3. Margin of safety
Intrinsic value is an estimate, and estimates are wrong in ways that surprise their authors. So we insist on paying meaningfully less than our appraisal — the gap between price and estimated intrinsic value shown as "upside" on the watchlist. A wide margin of safety is not a way to earn more. It is a way to be wrong and survive.
The ledger principle
Verdicts published here are never edited and never deleted. Each one is stamped with its date and the price at the time, and its result is graded automatically by the pipeline as prices move — not curated after the fact. If a verdict ages badly, it stays on the ledger next to the ones that aged well. A track record you can prune is not a track record.
What the site's sections mean
Screens filter a universe of stocks on quality-and-value criteria; passing a screen is an invitation to read, not to buy. Analysis covers one company per day and ends with a verdict: circle (inside our circle of competence and attractive), watch (interesting, wrong price or open questions), or pass. The Watchlist is the standing ledger of businesses we admire and the prices we would pay — a hypothetical model portfolio, with no real money behind it. Sentinel logs events on watched names. Letters sum up the week.
Sources and accuracy
Figures are drawn from public filings and market data feeds, processed automatically. They may be delayed, incomplete, or wrong. When our numbers and the company's filings disagree, believe the filings — then tell us.
Contact
Corrections, questions, and disagreements are welcome at wbai@moatledger.com. Disagreements especially — they are how a ledger stays honest.
"It is better to be approximately right than precisely wrong." We try to be approximately right, and to say clearly when we might not be.