Actionable insights — 10 Stocks for Life

1. Look for the control pair inside any thematic list, and let it size the theme's actual power

1. Scan the list for two names in the same industry, of similar age, given nearly the same argument.
2. Compare their realised returns over the same window. Any gap is the part the theme does not explain.
3. Attribute that residual explicitly — capital allocation, pipeline productivity, the starting valuation — and ask which of those you can actually assess in advance.
4. Downgrade the theme from a reason to buy to a reason to look, in proportion to the size of the gap.

• Lists that avoid publishing per-name returns, which removes the ability to run this test at all.
• Your own tendency to remember the 17,000% and forget the 9.8% from the same page.

2. Verify a longevity claim against the corporate lineage, not the founding anecdote

1. For any "founded in 18xx" claim, find out what the entity you would actually buy today is, legally and operationally.
2. Note mergers, spin-offs, bankruptcies and renamings. Each one breaks continuity in a different way and only some of them matter.
3. Ask what actually persisted through the break — the customers, the network, the licence, the brand — and treat that, not the date, as the durable asset.
4. Separate the founding date from the listing date: a business that was private for a century has an unobserved record.

• Founding dates that belong to an acquired predecessor rather than the listed parent.
• Long private histories: the founding-to-IPO gaps here run from 24 years (Sherwin-Williams was public within a century) to 124 (Colgate) — the unobserved period tells you nothing about shareholder outcomes.

3. Distinguish an asset that cannot be rebuilt from one that is merely expensive to rebuild

1. Ask the question directly: with unlimited capital, could a competitor recreate this asset today?
2. If the barrier is money, it is a cost advantage and it erodes when capital is cheap or when a rival is subsidised.
3. If the barrier is legal, geographic or political — a right of way, a licence, a planning regime, an accumulated trust position — it is genuinely irreplaceable and deserves a longer forecast horizon.
4. Rank the candidates by which kind of barrier they have, and be sceptical of any moat described only as "scale."

• Technological bypass rather than direct replication — nobody will lay parallel track, but freight can move by road or by pipeline.
• Regulated irreplaceable assets attracting regulated returns, which caps the upside the moat appears to promise.

4. Use purchase frequency and unit size as a forecastability test

1. For a consumer business, characterise the typical purchase: how small, how frequent, how habitual, how deferrable.
2. Small, frequent, habitual and non-deferrable purchases produce revenue that survives recessions almost unchanged — which is what makes long-horizon cash-flow forecasts defensible at all.
3. Test the opposite direction too: can the customer trade down to a cheaper substitute without effort? Habit is only a moat while switching feels like a decision.
4. Prefer this test to brand-strength language, which is unfalsifiable.

• Private-label substitution in exactly these categories — the habit belongs to the product, not always to the brand.
• Categories where a drug, a regulation or a fashion changes the underlying behaviour, which is the one thing this test assumes cannot happen.

5. Publish the caveat with the list, in the same document

1. When you produce a screen output, state explicitly what the screen does and does not claim.
2. Say whether a price has been considered. If none has, say so — a list of businesses is not a list of investments.
3. Give the reader the intended use: idea generation, watch list, or action.
4. Keep the caveat in the same document as the names, because the names travel and the caveat does not.

• The caveat appearing only at the end, after ten enthusiastic entries — placement determines whether it is read.
• Screens circulated as recommendations once the framing is stripped away, which is what happens to every list of tickers eventually.