The three-condition pricing-power test, how to verify an acquisition record statistically, and the discipline of naming the price at which a pass becomes a purchase.
1. Test pricing power with three conditions, not with brand or margin
The repeatable method
- Is the product mission-critical — can the customer operate without it? (Better still: is it legally required?)
- Is there almost no competition for this specific item, and what stops a new entrant?
- Does the product represent a trivial share of the customer's total cost? Compute the actual ratio rather than asserting it.
- Only when all three hold does an annual price increase go through unopposed. Two out of three is not enough.
Here: AmSafe — a legally required seatbelt, a 95%+ share, and ~$20,000 of seatbelts on a $30–50m aircraft, "less than 0.1% of the aircraft's overall cost." Hence 5–6% price rises "every single year."
Watch for
- The share-of-cost ratio drifting up. Repeated price rises eventually make the item visible in the customer's budget — that is the mechanism's own expiry date.
2. Find behavioural proof that the customer is price-insensitive
The repeatable method
- Do not infer insensitivity from margins — margins are the result you are trying to explain.
- Look for customer behaviour that only makes sense if price is irrelevant: not pursuing refunds, not tendering, not switching after an increase.
- Compare that behavioural threshold with the company's actual average selling price.
- If the threshold is above the price, price increases have room to run.
Here: "When a part costs under $800 and doesn't work on arrival, Airbus often doesn't bother requesting a refund. Instead, they simply order a replacement. The average price across TransDigm's product portfolio is around $1,000."
Watch for
- Procurement reform at a large customer. The anecdote describes a habit, and habits are exactly what a cost-cutting drive targets.
3. Check a serial acquirer's record against the base rate before believing it
The repeatable method
- Establish the base rate: roughly 70% of acquisitions destroy shareholder value.
- Count the deals and ask what the claimed hit rate implies. A perfect record across dozens of deals is statistically extraordinary — 0.3^82 by chance.
- Treat that as a prompt to verify rather than to admire: who computed the returns, on what definition, and was it audited by a credible outsider?
- Look for an independent corroborating signal — an award judged by peers, a competitor's testimony, or a third-party analysis.
Here: Howley's 50–82 acquisitions "all… created private equity-like returns (>15%)," corroborated by the Singleton Prize (judged by Will Thorndike, Todd Combs and Mark Leonard) and by the acquired McKechnie CEO's testimony — "every day we were trying to copy them."
Watch for
- Self-reported deal returns. The figure comes from management and its own advocate; the corroboration is qualitative, not an audit.
4. Weight the testimony of a competitor you acquired above any other source
The repeatable method
- Look for cases where a direct rival came inside the business — through acquisition, a hire, or a partnership.
- Their assessment is the least contaminated evidence available: they had every incentive to believe the model was replicable, and they had tried.
- Weight a reversal of view (they tried to copy it and could not) more heavily than any management presentation.
Here: the former CEO of McKechnie Aerospace, a self-described "TransDigm 2.0" bought for $1.3bn: "These guys are so much better than I ever imagined. I totally underestimated how they do what they do."
Watch for
- The obvious selection effect — you rarely hear from the rivals who succeeded, because they were not acquired.
5. Separate razor from razorblade, and value them differently
The repeatable method
- Split revenue into the original sale and the replacement stream, and get the margin on each.
- Estimate the installed life of the product — the longer it is, the larger the aftermarket relative to the original sale.
- Ask why nobody enters the replacement market: in shrinking, ageing fleets the prize is falling while the approval cost is not, so entry is unattractive by construction.
- Value the aftermarket as recurring revenue and the original sale as cyclical, not as one blended business.
Here: Defense 43% / Commercial OEM 25% / Commercial Aftermarket 32%, with aftermarket margins "much higher"; aircraft fly 30–50 years; and "80% of its revenue comes from products for which it is the only provider."
Watch for
- A regulatory route around the monopoly. PMA approvals let a third party make an approved replacement, and their use "has increased in recent years."
6. Treat approval cost as a moat — and measure it by how long incumbents have survived
The repeatable method
- Identify the regulatory approval a new entrant would need, and estimate its cost and elapsed time.
- Test the barrier empirically: how old are the incumbents in this niche? Longevity is the observable proxy for entry difficulty.
- Prefer the boring components: the more mundane and less performance-differentiating the part, the less anyone bothers to attack it.
- Check whether a regulator could create a bypass — that is the barrier's single point of failure.
Here: "Even the soap dispensers on planes need approval from the Federal Aviation Administration"; "most of the airplane parts manufacturers have been in business for over 70 years"; and the industry saying, "airlines buy new planes for fuel efficiency, not seatbelts."
Watch for
- The bypass being the same regulator. PMA is an FAA programme — the body that built the barrier also runs the route around it.
7. Convert a pass into a standing order by naming the price
The repeatable method
- Complete the full quality work even when you expect to decline — it is reusable the moment the price moves.
- State the reasons for the pass in order of importance, separating structural objections (size, leverage) from price.
- Convert the price objection into a specific multiple, then into a specific share price, and state the distance from today.
- Record it so the decision is triggered by the market rather than remade under pressure.
Here: Total Quality Score 8.0/10 and "we are not buying" on size, gearing and valuation — followed by "we would start to become interested at a FWD PE of 25x… a stock price of $1,010. (12.5% below today's stock price)."
Watch for
- The article's own counter-argument: TransDigm "has always been expensive but kept generating shareholder value." A multiple-based trigger can keep you out of a business whose multiple never mean-reverts.
8. Read a negative ROE before treating it as a failure
The repeatable method
- When return on equity is negative, check whether the denominator is negative rather than the numerator.
- Negative equity from debt-funded buybacks and special dividends is a capital-structure choice, not unprofitability.
- Switch to return on invested capital, which is unaffected by the distinction, and judge on that.
- Then judge the leverage on its own terms — interest coverage and net debt to EBITDA — rather than through the ROE artefact.
Here: ROE −26.1% because "their capital structure relies on Debt instead of Equity" and a large special dividend cut retained earnings; ROIC is 17.0%. The leverage is then failed separately and explicitly: 2.5x interest coverage, 5.9x Net Debt/EBITDA, 46.6% goodwill.
Watch for
- Debt raised to fund a distribution rather than an acquisition — flagged here as the one management decision that "doesn't make much sense in our opinion."
Methods distilled from the archived Compounding Quality post for personal study. Not investment advice.