Actionable insights — 20 Timeless Stocks to Own

1. Use age as a first-pass survival screen, and justify it with the base rate

1. Start from the failure base rate, not from the survivors: roughly half of all businesses are gone in five years and four-fifths in ten.
2. Filter a universe by continuous corporate age — 100 years is the threshold used here — and treat what remains as having passed a test no financial statement can replicate.
3. Read the age as evidence of adaptability, not of stability: ask what the business abandoned in order to survive.
4. Apply Lindy only to things that do not age biologically — institutions, standards, brands, networks. It says nothing about a person, a patent, or a product.
5. Treat the output as a candidate universe, then run your normal quality and valuation work on it.

• Survivorship bias in the supporting evidence: a portfolio of companies selected today for being 100 years old could not have been assembled in 2000, which is when the cited backtest starts.
• Legal continuity mistaken for business continuity — a name that survived through bankruptcy, bailout or merger is not the same entity that started.

2. Convert expected corporate lifespan into an explicit terminal-value argument

1. State the claim in discounting terms: value equals all future cash flows discounted to today, so a longer expected life mechanically adds value.
2. Estimate, however roughly, how many more years of cash flow you are underwriting — and notice how much of a typical valuation sits beyond year ten.
3. Where the business is genuinely Lindy, allow yourself a longer explicit forecast horizon; where it is not, cut the horizon rather than the growth rate.
4. Sanity-check the other direction: what would have to be true for the business to be gone in fifteen years? If the answer is nothing you can name, the age argument is doing real work.

• Terminal value doing all the work — an argument that can only be settled decades from now cannot be falsified now.
• Discount rates: at any realistic rate, cash flows beyond thirty years contribute very little, so the Lindy premium is smaller than it feels.

3. Separate "survived" from "compounded" before treating the screen as a buy list

1. For every name the age screen produces, pull the ten-year return alongside the multi-decade one.
2. Where the two disagree sharply, the business has been durable but the shares have not — usually a sign the price already embedded the durability.
3. Require a second, independent reason to own it: a price, a re-rating, an operational turn. Age alone is a reason to study, never a reason to buy.
4. Do the same for the list as a whole: count how many of the names produced their return before your holding period would have started.

• A list published without prices being read as a list of buys — the same names carry hard fair values in this archive's portfolio updates and none is given here.
• Total-return figures measured from 1990, which flatters anything that was cheap in 1990 regardless of the business.

4. Classify durability into three testable kinds rather than accepting "great brand"

1. Ask first whether the underlying need is permanent — food, shelter, health, energy, payment, safekeeping, story. If the need can disappear, nothing else matters.
2. Then ask whether the company owns something physically or institutionally irreplaceable — a network, an installed base, a licence, a trust position.
3. Then ask whether the economic model has migrated away from the capital-hungry part of the industry towards fees, rent or royalties.
4. Score each name on all three. One alone is weak; the durable compounders on this list score on at least two.

• A permanent need served by a replaceable supplier — the need for energy does not guarantee any particular energy company.
• Fee-based models whose fee is set by the party paying it: franchisees and hotel owners renegotiate.

5. Read the omissions in a durability list as carefully as the arguments

1. For each name on any published list, write down the one obvious current controversy — the litigation, the demand shock, the disrupted segment.
2. Check whether the author addressed it. Silence on a live issue is information about the format, not about the risk.
3. Distinguish long-horizon framing (which legitimately ignores this quarter) from selective omission (which ignores a threat to the terminal value itself).
4. Only the second is disqualifying — but you have to name it yourself, because the list will not.

• Format-driven optimism: a "timeless businesses" piece has no natural place to put a risk, which is precisely why the reader must supply one.
• Your own reluctance to spoil an appealing story — the omissions are always the comfortable ones.