Figures the watch industry repeats, and what happened when we tried to trace them. Untraced does not mean false. It means nobody can show where the number came from — which is a different thing, and a checkable one.
59 claims · 1 disproved · 5 contested · 23 untraced · 30 traced
There is no legal definition, no industry threshold, no register, no certification and no body that adjudicates a claim. Compare "Swiss made", which is protected in ordinance. "In-house" cannot be breached because there is nothing to breach.
That figure, printed on Tudor's own certification page among others, is METAS-N001's tolerance for category 1a movements only (fitting diameter over 26mm). The same table sets 0 to +6 seconds a day for category 1b (20–26mm) and 0 to +7 for category 2 (20mm or under) — a three-step scale collapsed into one number by most of what gets published about it, brand pages included.
METAS's own N001 standard defines the "testing laboratory" — the entity that performs the tests — as one that "can be the applicant's in-house laboratory or the laboratory of a third-party entity commissioned by the applicant," acting under the applicant's own responsibility. METAS's stated role is to monitor that laboratory through statistical analysis of its results and periodic sample checks, not to run the day-to-day testing itself.
Depends entirely on what "shock" means. Some manufacturer language claims better rate-keeping through a shock event for specific designs; separately, and better documented, silicon is a more brittle material than the metal alloys it replaces and cannot be bent back once deformed. The two claims get run together in coverage and describe different things.
It does not, where the seller is a trader. The Consumer Rights Act 2015 attaches to the contract between trader and consumer and says nothing about authorised distribution, so the statutory rights at a grey dealer and an authorised dealer are the same ones. What differs is the manufacturer's voluntary guarantee, which is a separate thing sitting on top.
Against the January list it cannot have: the metal moved $3.10 a gram between the two lists, and recovering $2,200 at that rate needs 710 grams of fine gold in the watch. Against an autumn 2025 baseline it comfortably can. Rolex publishes neither the baseline it prices from nor the weight of gold in the watch, so the claim cannot be tested from outside — which is a different objection from saying it is false.
Records are quoted constantly in the trade. A specific result needs a specific lot reference, and the sales in question are not all documented in a form a reader can be pointed at, so the piece cites the existence of specialist watch departments rather than a hammer price. The direction of the reappraisal is not in dispute; a number attached to it would need a citation we do not have.
No regulator has tested that question for oscillating components. The ETA restriction is documented and was wound down under a competition settlement; nothing establishes the same for the regulating organ, and this publication does not assert it.
The figure the trade uses and the one Rexhepi's own comments about staying small are consistent with, but no audited or filed production number exists for a private Geneva atelier. Treated as the working assumption it is.
We have found no published rate whose method and cohort are stated. Every figure we have seen is asserted. If one exists with a traceable denominator, we will correct this and write the original piece.
The figure circulates in trade coverage and in this masthead's own commissioning brief. We could not trace it to a published policy, a service contract, a filing or an authority's finding. COMCO's 2018 report records that independent watchmakers told it the required tooling and premises exceeded their means, and the General Court calls the investments "large" — neither puts a number on it, and the service contracts are not public.
Article 8 makes any breach subject to the criminal provisions of the Trade Mark Protection Act. We could find no published count of prosecutions or convictions. FH's anti-counterfeiting pages report seizures of fakes — a different offence — and the most recent figures there date from 2014. A rule is only as strong as its enforcement record, and this one's is not public.
No filing, no regulator's finding, no company statement. It appears in trade coverage without attribution, gets repeated, and hardens into fact. Swatch Group's own page calls the company "the leading Swiss specialist" in oscillating and escapement parts, which is documented and sufficient; the percentage is not.
Repeated across trade coverage and attributed to Rolex, but neither of Rolex's own current pages for its two hairsprings — Parachrom (a metal alloy) and Syloxi (silicon) — states a number; both use only qualitative language. No dated Rolex statement carrying this figure could be located.
There is not one on the record. Reuters asked and reported that "Rolex declined to comment for this story", building the report instead on two luxury research platforms and two dealers; the trade coverage that followed carries no company comment either. Every published explanation of why gold prices rose is inference by people outside the company.
Establishing it needs component-level disclosure from every candidate, and none of them publish it. What is checkable is what Seiko demonstrably makes, which is a great deal; the superlative is not, and the piece describes rather than ranks.
Real in the trade, and the pitch for the piece raised it. But the Prospex, Presage and King Seiko pages resolve identically across the company's own global, US, UK and Japanese paths, so it could not be evidenced from the record and was not printed as a finding.
Attributed on record to Jason Champion, AWCI's director of education, in trade coverage — a named source, not an invented one — but AWCI has not published the ratio itself or the survey it would come from, so it cannot be checked.
The date is repeated almost everywhere the watch is written about. It does not appear on Yema's own history page, which does not mention 1963 at all — the brand's Superman entry is dated 1967 and concerns the bezel-lock patent, and its collection page says only "first introduced in the 1960s". A brand that documents its own timeline in detail and omits the founding date of its most famous model is worth noticing.
No figure is published by the company, and the estimates that circulate are ones it says deviate on average by 24%, with four of twelve brands more than 40% off. The corrections it has issued are partial — a margin here, a retail share there — and cannot be assembled into a revenue line. Nobody outside the company knows, and the company has chosen not to say.
Same trade-press origin as the figure above, repeated without a citation across multiple outlets; not published anywhere on awci.com.
The General Court refers at paragraph 112 to "the reduction in the number of independent repairers affiliated to a national association of independent repairers" without quantifying it. CEAHR's own evidence put the fall in association membership in Germany and the UK at 15-20% between 2004 and 2010, and acknowledged that this measures membership rather than the workforce, since joining is not compulsory in most member states. No authority, trade body or register publishes a time series of independent watchmakers in business.
No public figure exists that we can trace to a source. This is the denominator the commissioned piece needed, and its absence is why that piece is not this one.
Rolex publishes no case, bracelet or gross weight for any reference, in its own specifications or anywhere else. Dealers weigh individual watches and publish the result — one listed ref. 18238 at 129.4 grams gross — but an individual second-hand watch on a dealer's scale is not a factory figure, and no factory figure exists to check it against.
Circulates widely in trade coverage, attributed vaguely to "AWCI records" and "the Bureau of Labor Statistics." AWCI's own site publishes no such figures, and no specific BLS occupational series covering 1953 could be located.
Rolex, Tudor and Omega all refused every attempt to read their published warranty terms — 403 to a browser user-agent at rolex.com and tudorwatch.com, and no response at all from omegawatches.com. The terms are published, and we could not reach them to quote them, so this piece makes no claim about what any individual manufacturer will honour.
None of WatchCharts, Chrono24 or EveryWatch publishes transaction-level data. All three describe internal checks — WatchCharts' 30-day weighting, Chrono24's own-marketplace visibility, EveryWatch's "AI and manual processing" with "median effects" — but none allows an outside party to trace a specific published figure back to the sale it rests on, outside the auction-house portion of the market.
Widely repeated, including on Wikipedia. We have not found a filing, an audited figure or a company statement behind it. Plausible for the period and the plant; still nobody's number.
The premise of our own commission, and widely repeated. mb-microtec's pages say trigalight are "the world's smallest" gaseous tritium light sources, which is a claim about size, not share. We found no production or market-share figure from the company, a regulator or a filing.
The Financial Ombudsman Service: "A bank or lender doesn't have to raise a chargeback. But if valid reasons exist, it can be good practice for them to do so." Chargeback is a card scheme rule, not a statutory right, and the schemes' rules differ from each other.
Stated by traser in its own international warranty instructions, which also say the watch can be returned for disposal and give a Swiss Federal Office of Public Health approval number.
Reported by Forbes in its coverage of the sale, which the piece cites. Traced to named reporting of a public auction result rather than to trade repetition.
Federal Patent Court case S2018_006: CSEM sued Cendres+Métaux Microtech over the CMK1 movement used in Horage's Multiply watches; the court found the patent valid and granted interim injunctive relief against the movement maker.
Ball's own FAQ. Ball does not say how the comparison was measured, and we have not tested it; it is traced to the brand, not verified.
Chrono24's own methodology note, published with its March 2026 index update, on about.chrono24.com.
JCK's 3 August 2026 report on the EveryWatch H1 2026 Secondary Market Report, corroborated by The Idea Farm's syndicated summary of the same report.
JCK's 3 August 2026 report on the EveryWatch H1 2026 Secondary Market Report.
Both read from the LBMA PM benchmark daily series. Rebased to the 2 January setting of $4,352.95, the July low is 8.3% below where the year began, and all fifty-one settings from 1 June to 10 August sit below the 5.0% the yellow gold list rose.
It has more than doubled. The PM benchmark was $1,920.90 on 10 August 2023 and $4,335.55 on 7 August 2026, a rise of 125.7%. The trade's version of this claim is the one part of the explanation that survives checking.
Both figures are the LBMA PM benchmark, read from the LBMA's own published daily series. The fall between the two is 17.7%.
Net sales by geographical region, 2025 annual report. A fall of CHF 368m, disclosed to the million, in the same document that gives brands adjectives.
Tony Traina's lot-by-lot analysis on Unpolished, relayed with the brand-level figures by SJX Watches; Traina's own newsletter is paywalled beyond the aggregate figures and methodology note, which were read directly.
Luminox's own Light Technology page, read 10 October 2026. The page does not use the word radioactive.
Its reply of 5 March 2026, in those words.
The Financial Ombudsman Service, in its own words: "it's the cash price of the goods or services that matters, not what you paid on your credit card or loan – for example, Section 75 applies even if you only made part of the payment using credit".
Section 75(3)(b) of the Consumer Credit Act 1974, in terms. The Financial Ombudsman Service states the same limits and adds that it is the cash price that matters, not the amount put on the card.
Read off the segment information notes in the half-year report and the 2025 annual report. Corporate is shown separately and is explicitly not a segment.
Published in its open letter of 27 February 2026, rebutting the claim that Longines had become loss-making. It appears nowhere in the annual report.
COMCO Secretariat, final report of 8 May 2018, case 32-0256, paragraphs 189-190. The finding and the decision not to act on it are in the same document, on consecutive pages.
Read off Yema's own Superman collection page on 5 August 2026.
Article 1 paragraph 3 of the ordinance removes the component that enables the watch to be worn from the definition of a watch. FH's guide states the consequence in terms, and adds that even a bracelet soldered on permanently keeps its costs out of the sum.
EP1422436B1 (CSEM, filed 2002) expired 25 November 2022. A narrower patent on a different crystal orientation, held by Master Dynamic Ltd and filed in 2016, runs to 2036. The Swiss Federal Patent Court upheld the original patent against a rival design as recently as February 2019.
European Commission, Decision C(2014) 5462 final, Case AT.39097 "Watch Repair", footnote 168, which reproduces the progression from CEAHR's submission of 20 July 2005, Annex 11, and records that Italy appears in the same table with no data. The figures are not in the 2017 General Court judgment; the decision is where they are set out.
Both figures are read off the Federation of the Swiss Watch Industry's own worked examples on page 18 of its guide, which set out Variant 1 and Variant 2 side by side on identical components and identical costs.
Stated by FH's guide, which says the effect is to prevent a weakening of the Swiss made concept. This is the provision that most rebuts a cynical reading of the label and it is why the piece does not offer one.
Read off FH's own list, status 18 September 2025. The steel entries are numbered 18 and 19 at 90%; the titanium entries 22 and 23 at 95%.
CSEM's own 2002 patent states the figure directly: "the thermal drift of a spiral spring made of silicon is thus about 155 seconds/day... incompatible with watchmaking requirements which are of the order of 8 seconds/day." It is also why no commercial silicon hairspring ships as raw silicon — every one carries a compensating oxide layer.
The company's own "About our Price Guide" page, archived 24 April 2024: "We use the transaction price if it is known, or the last asking price otherwise. In cases where an insufficient number of transactions exist, we may also supplement the data with prices of unsold listings." The live page could not be independently re-checked; see the sourcing note in the text.
Stated by Yema itself, in those terms, on its in-house calibers page. Quoted rather than characterised, because the precision is the point.