A Part Recall Names the Supplier. Concentration Disclosures Decide Whether It Matters.

A recall notice is a manufacturing document with a legal timetable stapled to it. It states what failed, how many units are exposed, and in several regimes it names the firm that made the part. What it does not state is whether the firm that made the part now has a business problem or merely an incident. That second question is answered somewhere else entirely, in filings written months earlier, before anyone knew the part was bad.

The working frame here is narrow. The defect sets the size of the remediation. Concentration sets who absorbs it. Two suppliers can ship the identical failing component into the identical application and land in completely different financial positions, and the variable that separates them is usually disclosed — imperfectly, and on a lag — in documents that were public before the recall existed.

What the recall filing is required to contain

The reporting obligation is specific enough to be useful. Under 49 CFR 573.6(b), a motor vehicle or equipment manufacturer must submit a defect and noncompliance report not more than 5 working days after a defect has been determined to be safety related, or a noncompliance with a federal motor vehicle safety standard has been determined to exist. The clock starts at the determination, not at the press coverage.

The contents are itemized. Paragraph (c)(3) requires the total number of vehicles or items of equipment potentially containing the defect or noncompliance. Paragraph (c)(4) requires something different and more interesting: the percentage of vehicles or items of equipment ... estimated to actually contain the defect or noncompliance. Paragraph (c)(6) requires a chronology of the principal events behind the determination, including a summary of all warranty claims, field or service reports, and other information, with their dates of receipt. Paragraph (c)(8)(iv) reaches down the chain and calls for identification of the component manufacturer and, where applicable, the component's country of origin.

Other regimes run different clocks for the same category of event. Under 16 CFR 1115.14(e), a consumer product firm should report immediately, that is, within 24 hours after obtaining information reasonably supporting the conclusion that a product contains a defect which could create a substantial risk of injury. Paragraph (d) of the same section allows a reasonably expeditious investigation to evaluate reportability, and says that investigation should not exceed 10 days unless a firm can demonstrate that a longer period is reasonable. On the medical device side, 21 CFR 806.10(b) requires the correction or removal report within 10 working days of initiating the correction or removal.

Sources: 49 CFR 573.6(b) and 573.7; 16 CFR 1115.14(d) and (e); 21 CFR 806.10(b), eCFR text current as of 23 August 2026.

Three regimes, three clocks, and none of them interchangeable. A trader reading across sectors who assumes the automotive five-working-day rule applies to a consumer product or a device is working from the wrong calendar.

Two numbers that get collapsed into one

Coverage of a recall almost always leads with the (c)(3) population — the units potentially affected. That is the large number. The (c)(4) estimate — the share thought to carry the defect in fact — is the smaller number and rarely travels. They do different work. Inspection and notification effort scales with the exposed population; part replacement scales closer to the subset that fails inspection. Where the two diverge sharply, a headline built on (c)(3) and a cost estimate built on (c)(4) can differ by an order of magnitude, and both can be reported accurately.

There is a second, slower disclosure that gets almost no attention. Under 49 CFR 573.7, the campaign is reported quarterly for six consecutive quarters, beginning with the quarter the campaign was initiated. Those reports carry the number of units inspected and repaired, the number inspected and determined not to need repair, and the number unreachable for inspection through export, theft, scrapping or failure to receive notification. That is a public completion-rate series with an eighteen-month tail, which is a very different horizon from the one a recall headline implies.

Concentration is the variable, and it is disclosed early

The disclosure that decides whether a defect stays local sits in the annual report, not the recall filing. 17 CFR 229.101 — Item 101 of Regulation S-K — requires a registrant to describe, among other things, revenue-generating activities, products and/or services, and any dependence on revenue-generating activities, key products, services, product families or customers, including governmental customers, at paragraph (c)(1)(i). For smaller reporting companies, paragraph (h)(4)(vi) asks directly for dependence on one or a few major customers.

Here is the part worth pausing on. Item 101 as it currently reads contains no percentage threshold at all. There is no ten percent line, no fifteen percent line, nothing. The requirement is materiality-based, and the materiality judgment belongs to the filer. The Commission's 2020 modernization of Items 101, 103 and 105 took effect on 9 November 2020 and moved this disclosure toward a principles-based standard. A specific percentage appearing in a filing is therefore the filer's own disclosure choice, and comparing that percentage across two filers is comparing two judgments, not two readings of the same gauge.

Risk factors carry the rest. 17 CFR 229.105 requires a discussion of the material factors that make an investment in the registrant or offering speculative or risky, organized under relevant headings, with risks generic to any registrant pushed to the end under General Risk Factors. The section must be in plain English under 17 CFR 230.421(d). And there is one structural tell built into the rule: if the risk factor discussion exceeds 15 pages, the registrant must add a summary of no more than two pages in bulleted or numbered form at the front. The presence of that summary is itself an observable — it tells you the risk section crossed fifteen pages without your having to count.

Sources: 17 CFR 229.101(c)(1)(i) and (h)(4)(vi); 17 CFR 229.105; 49 CFR 573.6(c)(8)(iv). No percentage threshold appears in the text of 17 CFR 229.101.

Where the money surfaces

Remediation cost has to appear somewhere in the financial statements, and there are two places worth checking that do not require any interpretive leap.

The first is a schedule. 17 CFR 210.5-04 provides that the schedule prescribed by 17 CFR 210.12-09 shall be filed in support of valuation and qualifying accounts. Section 12-09 sets out a six-column table: Description; Balance at beginning of period; Additions charged to costs and expenses; Additions charged to other accounts; Deductions; Balance at end of period. Firms are told to list, by major classes, all valuation and qualifying accounts and reserves not included in specific schedules. Read as a rollforward, that table separates two things a single balance cannot: whether a reserve was topped up during the period, and whether it was drawn down. A reserve that stays flat while a campaign is running is telling you something different from one that is charged and then consumed.

One honest limit. A separate tabular reconciliation specifically for product warranty liability is standard practice in filings, but the accounting-standard paragraph that mandates it could not be confirmed against a publicly retrievable primary source while preparing this piece, so no citation for it is given here. What is verifiable from the rule text is the Regulation S-X schedule above.

The second is the current report. Form 8-K Item 2.05 is triggered when a board or authorized officers commit to an exit or disposal plan expected to result in material charges, and calls for an estimate of the total amount or range for each major type of cost, along with the portion that will result in future cash expenditures. Item 2.06 covers material impairments. Both contain the same escape hatch and the same follow-up obligation: if the registrant is unable in good faith to make an estimate at the time of filing, it must file an amended report within four business days after it makes that determination. An Item 2.05 or 2.06 filed without numbers is not the end of the disclosure. It is a scheduled second event.

What the aggregate record looks like

Source: U.S. Food and Drug Administration, openFDA Device Enforcement API (api.fda.gov/device/enforcement.json), records with center_classification_date from 2015-01-01 to 2024-12-31, counted by recalling_firm.exact; total records in window 28,391; retrieved 23 August 2026. Records are product-level, not distinct recall events. Firm names withheld deliberately.

The shape matters more than any single bar. Across roughly ten years of medical device recall enforcement records, the most-recalled firm accounts for 875 records and the tenth-ranked firm for 314, while the two-hundredth-ranked firm still shows 29. The ten largest contributors together account for 4,737 of 28,391 records — 16.7 percent. The rest is a long, populated tail.

Read conservatively, that says recall involvement is a routine condition of manufacturing at scale rather than a rare shock. If a market priced every recall as an existential event for the maker, it would be repricing manufacturers constantly. It does not. What appears to get priced is the subset where remediation collides with concentration — where the affected part runs through a single qualified line, or where one buyer represents enough revenue that a de-listing decision is a revenue event rather than a warranty event.

Two caveats on the data, both material. These are enforcement records at product level, not distinct recall events; a single campaign covering many product listings generates many records, so the counts overstate event frequency and the overstatement is not uniform across firms. And this is one sector. The device recall distribution is not evidence about automotive or consumer goods, only an illustration that recall activity has a head-and-tail shape at all.

What Would Invalidate This

The frame is a lens for one specific situation, and it fails in identifiable ways.

• The loss is qualification, not remediation. If a buyer responds to a defect by removing the supplier from future programs, the damage is forward revenue on a multi-year design cycle. Nothing in the recall filing measures that, and the concentration disclosure understates it, because the disclosed dependence describes revenue already booked.
• The supplier does not file. Private companies, and foreign subsidiaries whose parent files nothing with the Commission, produce no Item 101, no Item 105, no Schedule II. The recall filing may name them; the concentration data does not exist. The frame simply has no input.
• Contract allocates the cost invisibly. Indemnification and supply agreements between a finished-goods maker and a component maker are generally not public. Who ultimately pays for a campaign can be settled by a contract term no reader can see, and the public split of the charge may not follow the public assignment of fault.
• The concentration disclosure is stale. An annual report filed ten months before a recall describes a customer mix that may have changed. Treat the filing date as the age of the data and discount accordingly.
• The firm recalls constantly. For a manufacturer sitting near the head of the distribution above, one more campaign carries close to zero marginal information. The frame only has content where the base rate is low.
• The horizon does not match. The resolution mechanism in this frame is a six-quarter completion series plus at least one annual report. A position measured in days will expire long before any of that arrives, and holding it on this reasoning is holding it on the wrong reasoning.

Concrete Framework

1. Separate the two population numbers before anything else. Locate the primary filing — the Part 573 report, the CPSC notice, or the FDA correction and removal report. Record the (c)(3) total potentially affected and the (c)(4) estimated share genuinely affected as two distinct figures, and note which one the coverage used.
2. Establish who filed. Determine whether the reporting entity is the finished-goods maker or the component maker, and check 573.6(c)(8)(iv) for the identification of the component manufacturer and country of origin. This is the step that tells you which balance sheet is in scope.
3. Date your concentration data. Pull the most recent annual report of the entity in scope. Read Item 101(c)(1) for dependence on key products, services or customers, and Item 101(h)(4)(vi) if the filer is a smaller reporting company. Write down the filing date next to whatever you find. That date is the age of the input, and it is the largest source of error in this frame.
4. Scan Item 105 by structure first. Check whether a two-page bulleted summary sits at the front of the risk factors. If it does, the discussion exceeds fifteen pages. Read the headings and subcaptions before the prose, and ignore whatever sits under General Risk Factors for this purpose.
5. Find Schedule II. Note the beginning balance, additions charged to costs and expenses, deductions, and ending balance for any reserve class plausibly connected to the campaign. Compare the same lines to the prior two periods rather than reading a single year.
6. Set a watch on Items 2.05 and 2.06. If either is filed without a numeric estimate, calendar a follow-up: the amended report is due within four business days of the determination being made.
7. Size to the six-quarter window. The 573.7 completion reports run for six consecutive quarters. If the thesis depends on completion rates or unreachable-unit counts, the position has to survive that long or the thesis is decorative.
8. Write the exit before the entry. State in advance which of the invalidation conditions above, if observed, ends the thesis — most usefully, the appearance of a qualification loss rather than a remediation charge, since that is the case where every number in this framework is measuring the wrong thing.

Nothing here is a recommendation to buy or sell any security. Rule citations are to the text as in force on 23 August 2026 and can change.