A supply outage is an availability failure, not a quality failure — underwrite time-to-restore

Claim

The instinct on a shipment-stopping event is to size the disaster. That is the wrong question. The evidence says the revenue almost always comes back, and the thing to underwrite is time-to-restore.

The base rate

Case	Cause	Time unable to ship	Revenue outcome	Permanent share loss
Stryker Mar-2026	Cyber — ordering, manufacturing AND shipping down globally; formally declared material under Item 1.05	~3 weeks	Q1 organic +2.4% vs an 8.0–9.5% guide (~$335–425M); Q2 organic +9.0%, FY guide maintained then raised	None
Masimo Apr-2025	Cyber	~2 months to full normal	$0 — FY guide intact, the affected quarter still printed +8%	None
Artivion Nov-2024	Ransomware	~2.5 weeks	limited; FY24 revenue still +10%	None
MKS Feb-2023	Ransomware	weeks	$160M deferred, ~$150M recovered	No
Applied Materials Feb-2023	Supplier knock-on	weeks	$250M out of Q2, recovered the same quarter — CFO: "none of that demand is perishable"	No
Clorox Aug-2023	Ransomware	weeks	Q1 sales −20%, then +20% organic the next quarter	No (~90% back)
Merck 2017	NotPetya wiper	months	$260M (2017) + $200M (2018)	No — Gardasil +37% in 2018
Henry Schein Oct-2023	Ransomware, then re-encrypted	weeks	$350–400M in Q4; no catch-up quarter ever came	YES
Integra LifeSciences May-2023	Regulatory — endotoxin, FDA 483, warning letters	2 to 3.5 years	~$148M cumulative charges	YES — a $7.1M customer-relationship intangible written down
BD Alaris 2019–2023	Regulatory — needed a new 510(k)	3.7 years	~$400M/yr	YES — fell from #1 US infusion share to below Baxter

The splitting variable

Quality impugned → share transfers and stays transferred. Merely unavailable → revenue defers and returns. Integra and BD were regulatory shutdowns: the clinician had years to re-qualify a competitor's product, and the reason for the outage was itself a reason to doubt the device. Philips Respironics is the same family (sleep/respiratory €2.2bn → ~€1bn while ResMed went $3.2bn → $5.15bn). Stryker, Masimo, MKS and AMAT were availability failures measured in weeks, and every one of them recovered.

Henry Schein is the one cyber case that lost customers, and it is the exception that proves the rule about what is being shipped. It distributes fungible dental consumables to episodic small-practice buyers who can reorder from Patterson with one phone call — no retraining, no clinical risk. The loss was visible in the competitor's filings first: Patterson's dental consumables internal growth ran +2.9% → +6.3% (peak-outage quarter) → +3.7% while Henry Schein's dental merchandise was −11.3%, a ~17-point swing. Consolidated revenue went $12.65B (2022) → $12.67B (2024), +0.2% in two years despite ~$955M of acquisitions.

How to apply

1. Ask "how long until they ship again?" before anything else. Weeks means a deferral; quarters means a problem; years means a franchise. Nothing else in the headline is as predictive.
2. Classify the cause as availability or quality. IT outage, plant fire, hurricane, port closure → availability. Recall, FDA warning letter, safety signal, endotoxin, 510(k) revocation → quality. Only the second reliably transfers share.
3. Model it as a reweighting, not an EPS cut. For a weeks-long outage the one-time cash cost is usually trivial in PV terms — on BSX it worked out to ~$0.10/share probability-weighted, against a $3.06 two-week price move. The damage, if any, belongs in the exit multiple and the bear weight, because the mechanism is share transfer, not cost.
4. Check the fraction of the quarter already banked. An outage starting with 26 of 64 business days left caps the quarterly hit mechanically, before any judgment is applied.
5. Watch the competitor's filings, not the victim's. Henry Schein's share loss was provable from Patterson's numbers months before management conceded "some customers have not returned yet."
6. Insurance recovers out-of-pocket cost, not lost sales — ~15% (Henry Schein, $60M limit against a $350–400M hole) to ~78% (Masimo). Never net it against revenue.
7. A third outcome exists: demand destruction. Baxter's IV fluids came back after Hurricane Helene, but demand stayed 10–15% below pre-hurricane levels because hospitals learned to conserve. Applies to consumables far more than to implants.

The counter-case — when to weight the tail higher

The base rate assumes the incumbent is defending from strength. Stryker was attacked into a market with no active competitive incursion and recaptured everything. Weight the permanent-loss tail higher when the victim is already losing share, when a rival is mid-launch, or when substitution requires no capital — see [[pattern-console-lock-in-is-symmetric]]. On BSX (Aug-2026) this pushed the transfer probability to ~30% rather than the ~0% the base rate alone would imply.

What would falsify this

A cyberattack or logistics outage lasting under a quarter that produced documented, durable share loss at an incumbent with capital-equipment lock-in — i.e. Henry Schein's outcome in a physician-preference-item market rather than a fungible-consumables one.

Related

• [[pattern-console-lock-in-is-symmetric]] — the mechanism that decides which side of this a device maker lands on.
• [[pattern-cyber-incident-prices-on-quantification-not-disclosure]] — the market-behaviour twin.
• [[principle-down-a-lot-is-not-cheap]] — a shock that does not change the cash flows does not create value on its own.

History

• 2026-08-26 — written during /analyze BSX, after BSX disclosed a cyberattack that took down global order processing and shipping. Built from eleven cyber precedents plus the two non-cyber "could not ship" cases (Integra, BD Alaris) that actually answer the share-loss question.