Actionable insights — Why you should invest in boring companies

1. Screen for durability with three questions before opening a spreadsheet

1. Is it essential? Would the customer's operation actually stop, or break a rule, without this product? Nice-to-have fails.
2. Does it have a replacement? Not "is there a competitor" but "could the customer switch at acceptable cost and risk." Retraining, revalidation and migration are the real barriers.
3. Is the cash flow stable and predictable? Look for recurring consumption — consumables, service contracts, subscriptions — rather than repeat one-off sales.
4. Only names passing all three earn the financial work. The screen is deliberately fast because its job is to shorten the list.

• "Essential" confused with "large" — a big market with many substitutes fails question two, which is the one that actually protects margins.

2. Value an installed base, not a unit sale

1. Split revenue into the equipment sale and everything the equipment consumes afterwards — instruments, accessories, service, software.
2. Treat each unit sold as an annuity opening rather than a transaction closing: the metric that matters is procedures per installed system, not systems sold.
3. Check that the recurring part is genuinely tied to the machine (proprietary consumables, service the maker controls) rather than open to third parties.
4. Then ask what happens if unit sales stall — a strong installed base keeps compounding through a placement slowdown.

• Regulatory or antitrust pressure to open the consumable market — the annuity depends on the maker controlling what the machine consumes.

3. Publish the multiple even when you are not making a valuation call

1. Make the quality case with its own evidence: margins, returns on capital, balance sheet, long-run shareholder return.
2. State the current multiple in the same block of numbers, so the reader is never given quality without price.
3. If you are not taking a valuation view, say the pitch is about the business — do not let the omission read as endorsement.
4. Keep the entry discipline separate and explicit, as the archive does elsewhere (Cintas: a stated 25x forward entry; HEICO: a published pass at 53-57x).

• A quality pitch that becomes a purchase without a price step — this issue provides the numbers to notice that gap, not to close it.

4. Match the temperament to the strategy, and design for boredom

1. Accept that a portfolio of essential, predictable businesses will produce very little to do — that is the design, not a failure of it.
2. Separate the desire for action from the investment process; route it somewhere it cannot cost money.
3. Build a cadence (monthly adds, a monthly ratings sheet) so that activity is scheduled rather than triggered by news.

• A scheduled cadence turning into automatic buying — the calendar should set the timing, the analysis should still set the name.

5. Check a cited track record against your own previous citation of it

1. When you quote an investor's record, note what exactly is being measured — a live fund, a backtest of a formula, gross or net of fees — and over which window.
2. Keep the same figure across issues, or explain the difference when it changes.
3. Treat a backtested formula return and a realised fund return as different classes of evidence, because they are.

• Round, spectacular numbers repeated without a source — the more quotable the statistic, the less often it gets checked.