Actionable insights — The Next Berkshire Hathaway

1. Identify a withheld company from the facts the teaser must disclose to be persuasive

1. List every specific, checkable claim the piece makes — returns, multiples, share counts, assets, biography.
2. Prefer structural facts over adjectives: a named asset or an exact share-count change identifies a company; "world-class management" does not.
3. Cross-reference two or three independent facts. One can coincide; three cannot.
4. Then verify against a source outside the teaser, and treat the identification as provisional until you do.
5. Note what the teaser deliberately leaves out — that gap is usually where the weakness is.

• Confirmation bias in the matching — decide what would disqualify a candidate before you go looking.
• Teasers that describe a composite or an idealised version of the company; the disclosures are chosen to sell, not to inform.

2. Separate the thesis from the sales copy, and grade them separately

1. Split the piece into two columns: verifiable claims about the business, and claims about the offer or the author.
2. Score only the first column. The second tells you about incentives, not about the investment.
3. Check the first column for internal consistency; a number that contradicts itself in the same piece was not checked.
4. Ask whether the same case, written without a subscription attached, would use the same emphasis.

• "Potential to 5x from here" and similar unanchored multiples — no time horizon, no method, not falsifiable.
• Urgency framing ("this week only", a $344,000 cost to the author) as a reason to skip the analysis.

3. Underwrite the runway, not the record

1. Ask what size the company was when it produced its historical return, and what size it is now.
2. Convert the target growth into absolute money required each year at today's base.
3. Judge whether the markets it operates in are large enough to absorb that amount.
4. Discount the historical rate by the change in scale — a compounder that is ten times bigger does not repeat its own record.

• The gap between the historical rate quoted for excitement and the target rate quoted for credibility; the second is the one to underwrite.
• A five-year record that spans one favourable cycle — for an insurer with equity investments, that is a market cycle as much as a skill measure.

4. Check a borrowed authority before you accept the framework attached to it

1. When a checklist is attributed to a named book or author, verify that the book actually contains it.
2. Ask whether the framework is being lent credibility it did not earn.
3. Judge the checklist on its own merits — it may be good and still be misattributed.
4. Treat a misattribution as a signal about the care taken elsewhere in the piece.

• A house framework that is sound being weakened by a false pedigree — the list is fine; the citation is not.
• Checklists where "the company we're buying has them all" — a seven-of-seven claim is a marketing statement, not an assessment.

5. Record where an idea came from, and treat network ideas differently from screen ideas

1. Log the provenance of every idea: a screen, a filing, a conversation, a competitor's holding.
2. Ideas from people carry information the screen cannot see — and also carry the risk that you are borrowing conviction rather than building it.
3. Apply the same analytical standard regardless of source; the provenance decides how you found it, not how you underwrite it.
4. Track the hit rate by source over time. It is the cheapest process improvement available.

• Credentialing by association: the great-niece of a famous investor, with "a close relationship with Warren Buffett," is offered as a reason the idea is good.
• Ideas that arrive with a conclusion attached; the analysis afterwards tends to be confirmatory.

6. Hold a teaser to the publisher's own evidentiary standard

1. Establish what the publisher normally provides before recommending something — for this archive, a fair value, a forward multiple against history, and a reverse DCF.
2. Check whether the promotional piece meets that standard.
3. Where it does not, note what is substituted: a track record, an analogy, a management target, an insider purchase.
4. Decide whether the substitution is justified by the business type or whether the standard simply slipped for the sales issue.

• The same name appearing in a screen with modest numbers and in a promotion with dramatic ones; reconcile them before acting.
• Free promotional issues that circulate far more widely than the paid analysis, so the weaker case is the one most readers see.