Actionable insights — 10 Stocks To Own Forever

1. Run the coffee-can question before any valuation work

1. Frame the whole exercise as a purchase you cannot reverse: "buy great companies without having the intention to sell them over the next few years or even decades."
2. Ask one question of every candidate — "Would I want to own this for the next 50 years?" — and answer it before opening a spreadsheet.
3. Accept the consequence: if the answer is yes, entry valuation stops being the binding constraint; if the answer is no, no valuation makes it a coffee-can name.

• Any candidate whose case depends on a catalyst, a cycle turning, or a re-rating — that's a trade, not a coffee-can holding.

2. Test relevance, not growth — "why will it still exist in 50 years?"

1. For each name write the single sentence that explains why its product or service is still being bought in five decades.
2. Reject anything whose answer depends on a technology staying current; prefer answers rooted in human needs, regulation, physical infrastructure, or habit.
3. Keep the answer short enough to say out loud — if it needs three paragraphs of qualification, the durability isn't there.

• Businesses whose durability sentence secretly contains a forecast ("if it wins the AI race") rather than a constant.

3. Classify the source of durability — and diversify across the sources, not the sectors

1. Sort each candidate by why it endures: demographic tailwind, industry structure (oligopoly), brand/niche leadership, an acquisition machine, or float-plus-culture.
2. Build the list so no single durability mechanism carries it — a shock that breaks one (say, a roll-up losing its cost of capital) should leave the others untouched.
3. Note that this diversifies differently from sector labels: two "financials" can rely on completely different mechanisms.

• A list where six of ten names endure for the same reason — that's one bet wearing ten tickers.

4. Prefer industry structure you can see over quality you have to infer

1. Count the credible competitors. Two to four national players with heavy physical or regulatory barriers is the shape to look for.
2. Ask what a new entrant would have to build — a national terminal network, a century of claims data, a trained surgeon base — and how many years and dollars that takes.
3. Treat longevity itself as evidence: a firm that has survived since the 1930s has already been tested by conditions you cannot model.

• New entrants funded by cheap capital, and regulatory changes that lower the barrier — the two ways an oligopoly stops being one.

5. Look for the decentralized serial acquirer — the compounding machine hiding in a dull industry

1. Identify companies that grow chiefly by repeatedly buying small businesses in a fragmented industry, then leaving them to run independently.
2. Require the target market to be large and fragmented enough that the acquisition runway lasts decades, and the individual deals to be small enough that no single one can break the company.
3. Judge the model on repetition and reinvestment discipline rather than on any single acquisition.

• Deal size creeping up, returns on acquisitions falling, or centralisation replacing the decentralised model — the classic ways a roll-up stops compounding.

6. Value the float engine — insurance as leverage without the risk of leverage

1. Look for holdings whose insurance operations collect premiums today and pay claims years later, leaving a pool of investable capital between the two.
2. Check that underwriting is disciplined enough that the float is roughly free — otherwise it is expensive borrowed money, not an advantage.
3. Then judge the investing side separately: float only compounds if the people deploying it allocate capital well.

• Reserve strengthening or a deteriorating combined ratio — signs the float is being bought rather than earned.

7. Underwrite culture and time horizon as a durable asset

1. Read management's own language for the horizon it plans on, and check that incentives and history match the words.
2. Prefer businesses where the culture is the moat's carrier — the thing that keeps the acquisition discipline, the underwriting discipline, or the brand control intact after the current CEO leaves.
3. Stress-test the succession explicitly: if the founder or figurehead were gone tomorrow, does the argument still hold?

• Incentive plans tied to revenue or size rather than returns, and prestige M&A — the two tells that the horizon has shortened.

8. Let the holding period absorb the entry multiple — but only for names that pass every other test

1. Recognise the arithmetic: over a multi-decade hold, compounding of business value dominates the entry multiple, so a fair price for a great business beats a great price for a fair one.
2. Apply the waiver only after the durability test has been passed — it is a licence to pay up for permanence, never a licence to skip diligence.
3. Say the concession out loud rather than rationalising the multiple away, so the trade-off stays visible.

• Using the 50-year framing to excuse a price on a business you would not actually hold through a 50% drawdown — the test of whether the coffee-can label was honest.