MKSI

PARTIAL :: earnings event :: tier RESOLVED

Registered claim

MKSI Q2 2026 (reports 2026-08-05) discloses AI / Electronics & Packaging chemistry revenue at or above 12% of total revenue (extending the 5%-to-10% mix doubling rather than plateauing at 10%) AND prints revenue at least 2% above the frozen Street consensus snapshot.

Registered 2026-07-28.

Resolution spec

Spec registered 2026-07-28.

• [A] CORE :: At the 2026-08-05 Q2 2026 print, MKSI discloses in primary-source materials that AI / Electronics & Packaging chemistry revenue is at or above 12% of total revenue for the quarter.
  kind: dated binary event :: event date 2026-08-05 :: evidence: primary source
• [B] confirm :: MKSI Q2 2026 total revenue is at least 2% above the frozen Street consensus snapshot.
  kind: numeric threshold :: q2-2026-total-revenue-vs-consensus-pct >= 2
  baseline 1210 (consensus snapshot, frozen 2026-07-31)
  note: Zacks Consensus Estimate for MKS (MKSI) quarter ended June 2026 revenues, $1.21B, via Zacks earnings preview republished on Yahoo Finance 2026-07-30; fetched 2026-07-31

Verdict rule: HIT if leg A and leg B. PARTIAL if both leg A and not leg B, or both not leg A and leg B. MISS if not leg A and not leg B.

As frozen: HIT = A && B; PARTIAL = (A && !B) || (!A && B); MISS = !A && !B

Tier history

• 2026-07-28: MEDIUM (initial). Trigger: First reasoning-based assessment on a prediction seeded today. Today's MKSI datapoint is genuinely supporting: an independent fund with capital at risk named MKS its single largest quarterly contributor specifically for 'enabling modern multi-layer chip architectures', which is the content-per-wafer mechanism behind the mix claim. But leg A asks for something narrower - that MKS DISCLOSE AI/Electronics-and-Packaging chemistry revenue at or above 12% of total. The mix has doubled 5% to 10% and E&P chemistry grew +22% YoY against 13% company growth, so 12% is a derived rather than invented threshold; the risk is that the ratio is a quotient whose denominator rises with a strong WFE quarter, and that MKS may simply not break the figure out.

Evaluated daily since registration: 18 evaluations. The 1 that moved the tier is listed above.

Resolution receipt

Verdict: PARTIAL. Resolved 2026-08-07.

Rule: HIT if leg A and leg B. PARTIAL if both leg A and not leg B, or both not leg A and leg B. MISS if not leg A and not leg B.

As frozen: HIT = A && B; PARTIAL = (A && !B) || (!A && B); MISS = !A && !B

• [A] FAIL: Result UNCHANGED (fail); detail CORRECTED 2026-08-09. MKS did disclose an AI chemistry mix figure with the Q2 print (release 2026-08-05, call 2026-08-06) - CEO John Lee, Q2 call Q&A: AI chemistry is 'incrementally better... think about 15% to 20% as the right number now for chemistry as a percentage of our chemistry for AI', up from 15% at Q1. The leg nonetheless fails because it is denominated differently from the issuer's metric: it requires AI / Electronics & Packaging chemistry revenue at or above 12% of TOTAL revenue, and MKS reports AI as a share of CHEMISTRY revenue only. At E&P of $381M on $1.25B total, with chemistry a subset of E&P, 15-20% of chemistry is roughly 2-3% of total revenue - far below the 12%-of-total bar, which MKS has never reported against and could not have met. The original detail's characterisation that 'the specific mix figure is not in the record' is superseded: the figure is in the record, on a different denominator.
• [B] PASS: actual_raw=1248 vs frozen baseline 1210 → 3.14% >= 2. Reported Q2 revenue $1.248B vs frozen snapshot $1210M.

Autopsy

Resolved PARTIAL on 2026-08-07. The core leg failed and the confirm leg passed, with revenue of $1.248B against a frozen baseline of $1,210M, 3.14% above it.

The first version of this autopsy was wrong about why the core leg failed, and the correction is the useful part. It recorded that the mix figure never appeared in the primary record, and that a company which had quantified the number before had this time chosen not to. That is not what happened. MKS did quantify it, on the 2026-08-06 second-quarter call, in reply to an analyst question. Chief executive John Lee said: "We had said, in '24, 5% AI chemistry as a percentage of our chemistry overall, then 10%, and this year 15% last quarter. I would say it's incrementally better. So think about 15% to 20% as the right number now for chemistry as a percentage of our chemistry for AI." The metric was re-disclosed without any obligation to do so, and it had gone up.

The leg failed on a denominator error made when the call was written, not on the company's silence. MKS reports AI chemistry as a share of chemistry revenue. This call asked for AI and Electronics & Packaging chemistry revenue at or above 12% of total revenue. Those are different quantities, and the company has never reported the second one. The magnitudes are not close: Electronics & Packaging revenue was $381M of $1.25B total, chemistry is a subset of that, and AI at 15% to 20% of chemistry lands near 2% to 3% of total company revenue. A 12%-of-total threshold was unreachable in this quarter and in every recent one, including the quarters the supporting evidence was drawn from. The leg could not have passed on any print.

The error arrived through the evidence rather than the arithmetic. The supporting evidence and the company-level record both stored the series as 5% to 10% of revenue mix. The actual series, stated by Lee on two consecutive calls, is 5% in 2024, then 10%, then 15%, now 15% to 20%, every figure as a percentage of chemistry revenue. The denominator was dropped when the evidence was first recorded, and the threshold was then built on the dropped denominator. Both records need repair.

For the thesis the correction runs the opposite way from the verdict. The claim that qualified materials capture share at process interfaces is directly reinforced by what Lee said: AI chemistry taking a larger share of chemistry revenue for a third consecutive step is precisely the mechanism that claim describes, stated at the company's own reporting granularity. That evidence had been scored as absent and so never entered the supporting set. It has since been recorded.

The lesson is not that a metric disclosed once need not be disclosed again, because MKS disclosed it again unprompted. It is that a threshold has to be denominated in the unit the issuer actually reports. Before a mix threshold is sealed, the denominator should be checked against the company's most recent statement of the metric rather than against an internal paraphrase of it.