Every plausible definition of SK Hynix capex intensity lands within 3 points of 28% — so the 11%-vs-mid-30% gap is not a definitional one, and capex/revenue should be demoted from the discipline threshold
The question
Verbatim: "Resolve the SK Hynix capex-intensity discrepancy (~11% Futurum vs 'mid-30% of revenue' guidance) by pinning the apples-to-apples number before using it as a discipline threshold."
Context: open follow-up #2 from [[2026-07-21-memory-maker-fab-financing-supply-side]], which carried both figures side by side and filed the gap as "different denominators/definitions." That diagnosis was wrong, and the real explanation is worse for the intended use: one figure is an unreconciled forward target and the other is an unsourced forward estimate, and neither is a measured ratio.
What we already know (from the vault)
- The two figures in tension, verbatim with their provenance. (1) "Futurum puts 2026 capex intensity at ~11% (vs Micron ~21%, Samsung ~25-30%, China's CXMT ~77%)"; (2) "SK Hynix guidance elsewhere is framed as 'mid-30% of revenue on a rolling 3-yr average'" — both in [[2026-07-21-memory-maker-fab-financing-supply-side]], sourced to a Futurum insight piece and a MarketScreener/Reuters headline respectively.
- SK Hynix became an SEC registrant in July 2026 (CIK 0002120882, SKHY/Nasdaq; F-1 2026-06-24, 424B4 2026-07-10), and the expectation was that IAS 16.73(e) would give a PP&E movement schedule with additions by class — better data than Micron's — flagged at the time as a build-time expectation, not a verified fact [[2026-09-06-vendor-capex-composition-memory-anchor]]. That expectation is now confirmed (section below).
- FY2026 capex guidance was raised to the "high 40 trillion won range" on the Q2 2026 call (2026-07-28/29), alongside explicit "CapEx Discipline" language [[2026-09-06-vendor-capex-composition-memory-anchor]].
- The vault has already been burned once by treating cross-issuer ratio levels as comparable. The DIO work found IFRS permits inventory write-down reversal while US GAAP forbids it, biasing Korean DIO levels above Micron's for non-cycle reasons; the resolution was "only each issuer's change against its own trailing floor is comparable" [[2026-09-07-memory-supplier-inventory-days-dio-anchor]]. The same constraint applies here and was not being applied.
- Standing doctrine is already "watch composition, not level" for memory capex [[2026-06-28-chip-memory-cycle-phase2-phase3-indicators]]. This brief supplies a second, independent reason for it.
What the web says
- The "mid-30%" figure is in the audited prospectus, verbatim — this is the strongest version of that claim and it supersedes the paywalled wire headline. "We maintain a disciplined capital allocation policy, targeting a capex-to-sales ratio in the mid-30% range based on a rolling three-year average. This approach is intended to preserve financial flexibility while supporting our long-term technology leadership." — Business → Our Strategy, item 3 ("Pursuing Production Capacity Expansion in Korea"), p.101. It is the only instance of a capex-ratio policy anywhere in the document. Critically, the filing nowhere defines the numerator or denominator of that ratio, states whether it is forward- or backward-looking, or reconciles it to any historical figure (SK hynix Form 424B4, 2026-07-10) — Tier A, audited/registered filing.
- Reported capital expenditure, audited, to the won. Consolidated statements of cash flows, investing activities, "Acquisitions of property, plant and equipment" (KRW millions): FY2023 8,325,138 · FY2024 15,945,534 · FY2025 27,518,924; Q1 2026 7,657,403 / Q1 2025 6,284,222. "Acquisitions of intangible assets" is a separate line: FY2023 454,710 · FY2024 717,106 · FY2025 1,060,419. The prospectus' own defined metric, footnoted, is "cash outflows for acquisition of property, plant and equipment" — i.e. it excludes intangibles (424B4, pp.16 and 62 and the audited statements) — Tier A.
- The IAS 16 accrual numbers exist and differ from cash capex. PP&E note additions: FY2025 30,172,818 (of which construction-in-progress 17,015,647 and machinery 11,867,710) and FY2024 17,955,821 — higher than cash capex by 2,653,894 and 2,010,287 respectively, unreconciled in the filing (an "Increase in other payables related to property, plant and equipment" line in the cash-flow note implies accrual-vs-cash timing). No FY2023 column (424B4, Note 12) — Tier A.
- Revenue, audited, plus a product-level disaggregation that yields a memory-only denominator. Revenue: FY2023 32,765,719 · FY2024 66,192,960 · FY2025 97,146,675. There is a single reportable segment ("the Group has a single reportable segment that is engaged in the manufacture and sale of semiconductor products"), but Note 24 splits revenue by product: FY2025 DRAM 74,904,134 + NAND 20,690,084 + Others 1,552,457. Solidigm is consolidated (SK hynix NAND Product Solutions Corp. at 97.48%, with a new Solidigm Inc. subsidiary established in Q1 2026), and the filing's own competitive charts state "SK hynix figures include Solidigm" (424B4) — Tier A.
- The prospectus contained no FY2026 capex figure at all. Its only forward statement is "In 2026, we plan to increase our capital expenditures considerably compared to 2025." The one 2026 number disclosed is project-specific: W7.9T of planned Yongin-fab-1 + P&T7 investment inside a W45.5T use-of-proceeds schedule running to 2030, plus W11.9T of EUV scanners for delivery by December 2027, against net offering proceeds of US$26.2B (≈ W39.9T at the filing's W1,523.5/US$ rate) (424B4, Use of Proceeds, p.51) — Tier A.
- Futurum's figure, fully provenanced. "Measured as capital expenditure to revenue, SK Hynix is the most capital-efficient of the majors — roughly 11% capex intensity in 2026 versus about 21% at Micron, 25-30% at Samsung's semiconductor arm (whose spend is inflated by a foundry), and a striking ~77% at China's state-backed CXMT in 2025." Brendan Burke, Futurum, 2026-07-10 — the same day the 424B4 was filed. The article cites no source for any of the four percentages, and it mixes periods inside one comparison (SK Hynix 2026, CXMT 2025, Micron and Samsung undated) (Futurum) — Tier C, no underlying data.
- Reported 2026 results, from the company. Q1 2026 revenue KRW 52.5763T; Q2 2026 revenue KRW 79.3187T, operating profit KRW 60.5426T (76% margin), H1 revenue past KRW 100T for the first time; end-Q2 cash KRW 88T, debt KRW 18.6T, net cash KRW 69.4T (SK hynix Q2 2026 release, 2026-07-29); FY2025 revenue KRW 97.1467T and FY2024 KRW 66.1930T (SK hynix FY25 release, 2026-01-28) — Tier A.
- Q2-2026 capex and the FY2021-2022 history are not yet primary. Quarterly capex Q2 2026 10,671,433 and FY2021 12,486,635 / FY2022 19,010,261 come from an aggregator (stockanalysis.com, KRX:000660, retrieved 2026-10-06) — Tier B. Validation: for every period the 424B4 also covers, the aggregator reproduces the audited capex and revenue lines to the won, which raises confidence in the Tier-B periods but does not certify them.
Convergences and contradictions
- The definitional axes are real and they are small. Computed on FY2025: cash PP&E / consolidated revenue 28.33%; cash PP&E + intangibles / consolidated revenue 29.42%; PP&E-note accrual additions / consolidated revenue 31.06%; cash PP&E / memory-only revenue 28.79%; accrual additions / memory-only revenue 31.56%. Total spread across every plausible definition: 3.24 percentage points. Currency contributes exactly zero (both numerator and denominator are reported KRW; the W1,523.5/US$ convenience rate cancels — FY2025 is $18.1B / $63.8B = the same 28.33%). Consolidation contributes nothing either, since Solidigm is inside both numerator and denominator. Definitions cannot explain a 24-point gap. They explain three points of it.
- Both headline figures use the same definition, and neither is a measurement. Futurum states its definition explicitly ("capital expenditure to revenue"); the company's policy is "capex-to-sales." Same ratio. The company figure is an unreconciled forward target that the filing itself never ties to an actual; the Futurum figure is an unsourced forward estimate published on a day when no FY2026 capex figure had been disclosed by anyone — the prospectus filed that same day says only "increase considerably." The gap is between two guesses about the same year, made six months apart, while the revenue forecast for that year roughly tripled.
- A contradiction the vault should record: SK Hynix has never actually run at mid-30%, on any definition, in any period the filing discloses. The rolling three-year 2023-2025 ratio — exactly the policy's own stated basis — is 26.41% on cash PP&E, 27.55% including intangibles, and 27.25% against memory-only revenue. Even the most generous single-year reading (accrual additions over memory-only revenue, FY2025) is 31.56%. The "mid-30% range" is a ceiling the company has been running several points under, not a description of behaviour.
Synthesis for RDCO
The defensible apples-to-apples series, named definition first. Definition: cash outflow for acquisitions of property, plant and equipment, from the consolidated statement of cash flows, divided by consolidated revenue for the same period, both in reported KRW, IFRS as issued by the IASB, fiscal year ending December 31, Solidigm consolidated, no FX conversion, no segment carve-out. This is the company's own footnoted definition of "capital expenditures," so it is the one that will stay reproducible against future 20-Fs.
| Period | Capex (KRW m) | Revenue (KRW m) | Capex / revenue | Capex / D&A | Capex / OCF | Tier |
|---|---|---|---|---|---|---|
| FY2021 | 12,486,635 | 42,997,792 | 29.04% | 1.21x | 63.1% | B |
| FY2022 | 19,010,261 | 44,621,568 | 42.60% | 1.37x | 128.6% | B |
| FY2023 | 8,325,138 | 32,765,719 | 25.41% | 0.63x | 194.6% | A |
| FY2024 | 15,945,534 | 66,192,960 | 24.09% | 1.27x | 53.5% | A |
| FY2025 | 27,518,924 | 97,146,675 | 28.33% | 1.98x | 51.6% | A |
| TTM to Jun-2026 | 35,232,021 | 189,170,615 | 18.62% | 2.37x | 27.8% | A/B mix |
| H1-2026 | 18,328,836 | 131,895,000 | 13.90% | 2.36x | 20.0% | A/B mix |
Arithmetic, checkable: 27,518,924 / 97,146,675 = 0.2833. 8,325,138 / 32,765,719 = 0.2541. (7,657,403 + 10,671,433) / (52,576,300 + 79,318,700) = 18,328,836 / 131,895,000 = 0.1390. Three-year window on summed flows: (8,325,138 + 15,945,534 + 27,518,924) / (32,765,719 + 66,192,960 + 97,146,675) = 51,789,596 / 196,105,354 = 0.2641. D&A denominators use the cash-flow-note depreciation + amortisation split (FY2025 13,099,311 + 830,819 = 13,930,130); note that the filing carries three non-tying D&A figures for FY2023 — 13,619,161 in the expenses-by-nature note, 13,282,991 as the cash-flow-note split, and a third in the aggregator series — and does not reconcile them, so capex/D&A should carry a stated numerator source. Confidence in the ratios given the inputs: high. Confidence in the FY2021-2022 and Q2-2026 inputs: moderate (Tier B).
Mapping each conflicting figure to a definition that reproduces it — the mapping is the answer, and one of the two does not map. "Mid-30%" reproduces only as a plan ratio: the later-guided FY2026 capex of ~KRW 48T is 35.0% of KRW 137.1T (48.0 / 0.35 = 137.1), and KRW 137T is ~1.41x FY2025 revenue — a plausible January-2026 forecast for FY2026. A second candidate: Q4-2025 standalone capex intensity was 36.16% (11,870,574 / 32,826,700), the most recently printed quarter when the policy was restated. Both land in the mid-30s; neither can be confirmed, because the filing refuses to define the ratio and the Reuters body is paywalled. "~11%" reproduces only as a forward FY2026E ratio, and only under assumptions no one disclosed: 48.0 / 0.11 implies KRW 436T of FY2026 revenue, which is 4.5x FY2025 and above even Q2-2026 annualised (79.3187 x 4 = KRW 317.3T); on a lower numerator it comes easier — 35.2 / 317.3 = 11.1%, or 40.0 / 363.0 = 11.0%. Verdict on the Futurum figure: not falsified, but not reproducible, and not a measurement. It is 2-3 points below the nearest actually-reported analogue (H1-2026 at 13.90%, Q2-2026 standalone at 13.45%), so it is directionally honest; it was nonetheless published before any FY2026 capex number existed in public disclosure, against an undisclosed revenue estimate, with no source cited for it or for the three vendors it is compared against. It should be carried in the vault as "Futurum estimate, 2026-07-10, inputs undisclosed," never as a figure, and it should not anchor a threshold.
Why capex/revenue has to be demoted, independent of which number is right. From FY2023 to the trailing twelve months, SK Hynix's absolute capex rose 4.23x (8.33T to 35.23T) while capex intensity fell from 25.41% to 18.62%. The metric moved the wrong way through the largest capex escalation in the company's history, because revenue moved 5.77x against capex's 4.23x. In a price-driven memory upcycle capex/revenue is not merely noisy, it is perversely signed: it falls fastest exactly when spending accelerates hardest, because ASP rather than volume drives the denominator. At quarterly frequency it is worse — 36.16% in Q4-2025 to 13.45% in Q2-2026, a 22-point swing in two quarters with no behavioural change that could justify it. A tripwire keyed to a capex-intensity threshold would have stayed green through 2026 and would stay green through a KRW 48T year. Report capex intensity as a ratio with its vintage attached; do not make it the threshold, and never compare two vintages of it.
The replacement, recommended explicitly, with its cost stated. Primary: absolute capex indexed to the prior-cycle peak, per maker. SK Hynix's prior peak was FY2022 at KRW 19.01T; FY2025 is 1.45x it, TTM 1.85x, and the guided FY2026 2.52x (48.0 / 19.01). That series says plainly what intensity conceals — this is the most aggressive absolute capex posture in company history. Companion, for normalisation: capex / depreciation and amortisation, 1.21x / 1.37x / 0.63x / 1.27x / 1.98x for FY2021-2025 and 2.37x at TTM, with a clean economic reading at 1.0x (replacing the asset base exactly); FY2023's 0.63x correctly marks the capex-cut phase that intensity's 25.41% completely missed. The reason to prefer it is denominator stability, not smoothness: over FY2021 to TTM, D&A moved 1.45x while revenue moved 5.77x, so movement in capex/D&A is attributable to the numerator — the behaviour being measured. Third, as a forward observable rather than a backward one, and the single best find in this brief: Note 29 contracted-but-unrecognised capital commitments, disclosed KRW 8,837,748m at 2024-12-31, 6,667,863m at 2025-12-31, and 30,921,877m at 2026-03-31 — a 4.64x jump, +KRW 24.3T, in one quarter, which is larger than any full year of cash capex before FY2024 and which leads cash capex by construction, since it is signed purchase orders not yet in PP&E. That is the discipline observable the intensity ratio was standing in for. The cost of switching, paid knowingly: (1) cross-vendor comparability on one scale is lost — the "SK Hynix 11% vs Micron 21% vs Samsung 25-30% vs CXMT 77%" line cannot be restated in capex/D&A without building a D&A series for all four, and for CXMT and Samsung's semiconductor arm that data may not exist; (2) capex/D&A levels are sensitive to depreciation policy and asset vintage, so cross-issuer levels are no more comparable than DIO levels were — same constraint, same resolution, each issuer against its own history; (3) prior-peak indexing requires a declared cycle-peak date per vendor, which is a judgement call where hindsight can leak in; (4) the commitments line needs a verified Micron and Samsung analogue before it can be a three-maker gauge. Net: these are per-issuer time series, not a cross-vendor leaderboard — a downgrade in what we can say and an upgrade in what we can trust.
One line for the risk register, separate from the measurement point. The guided KRW 48T requires H2-2026 capex of KRW 29.67T, up 62% on H1's KRW 18.33T — a steep back-load, and the KRW 30.9T of contracted commitments at 2026-03-31 is consistent with it landing. If it does, FY2026 capex/D&A sits near 2.9x (48.0 / ~16.5T, annualising Q2-2026 D&A of 4.03T) and absolute capex at 2.52x the prior-cycle peak, with "CapEx Discipline" still in the press-release headline and intensity printing in the low teens. That conjunction — record absolute spend, a falling intensity ratio, and discipline language — is the pattern worth watching, and the metric that was about to become the threshold is the one that would have hidden it. Separately: Futurum's Micron figure ("about 21%", year unstated) was not reproduced here and should not be relied on.
Why this is in the vault
It retires a specific number from the memory-cycle tripwire set. The Phase-2→Phase-3 indicator work in [[2026-06-28-chip-memory-cycle-phase2-phase3-indicators]] and the per-maker discipline gauge proposed in [[2026-07-21-memory-maker-fab-financing-supply-side]] were both heading toward a capex-intensity threshold, and this brief shows that threshold would have stayed green through a 4.2x capex escalation. It also specifies three concrete fields for the memory-vendor-capex/ anchor schema drafted in [[2026-09-06-vendor-capex-composition-memory-anchor]] — capex_vs_prior_cycle_peak, capex_over_da, contracted_capital_commitments — confirms that brief's unverified IAS 16.73(e) expectation (additions by asset class are disclosed, and they run 2.0-2.7T above cash capex), and downgrades the Futurum ~11% from fact to attributed estimate wherever the vault carries it.
Open follow-ups
- Compute Micron's capex/revenue, capex/D&A and capex-vs-prior-peak on this exact definition from its 10-K cash-flow statements, and test whether Futurum's "about 21%" is reproducible for any Micron fiscal year. Micron is the one vendor in that four-way comparison with fully-tagged US GAAP filings, so if the secondary figure fails to reproduce there, the whole cross-vendor intensity comparison can be dropped rather than merely caveated.
- Does contracted-but-unrecognised capital commitment disclosure lead cash capex reliably, and do Micron and Samsung disclose a comparable line? SK Hynix's jumped 4.64x in one quarter (KRW 6.7T → 30.9T) ahead of a capex guidance raise, which is the right shape for a leading indicator, but n=1 observation on one issuer. Micron's 10-K commitments note and Samsung's DART business report would establish whether this is a three-maker observable or an SK-Hynix-only artefact. (Distinct from the hyperscaler purchase-commitment disclosure already catalogued in [[2026-08-01-hyperscaler-capex-merchant-vs-inhouse-silicon-decomposition]], which is demand-side.)
Related
- [[2026-07-21-memory-maker-fab-financing-supply-side]] — the parent brief; carries both conflicting figures and filed this as open follow-up #2
- [[2026-09-06-vendor-capex-composition-memory-anchor]] — the capex anchor schema this brief adds three fields to; source of the "high 40 trillion won" guidance and the SEC-registrant status whose IAS 16.73(e) expectation is now confirmed
- [[2026-09-07-memory-supplier-inventory-days-dio-anchor]] — the precedent for "cross-issuer ratio levels are not comparable, only each issuer against its own history"
- [[2026-06-28-chip-memory-cycle-phase2-phase3-indicators]] — the indicator set whose capex leg this brief re-specifies; origin of "watch composition, not level"
- [[2026-07-15-capex-financing-layer-as-memory-cycle-phase-marker]] — the grandparent brief that rejected demand-side capex financing as a phase marker
- [[2026-09-26-financing-led-capital-cycle-phase3-precedent]] — the financing-led Phase 3 path whose F-block thresholds sit alongside this one
- [[2026-08-01-hyperscaler-capex-merchant-vs-inhouse-silicon-decomposition]] — the demand-side purchase-commitment disclosure catalogue, the contrast case for follow-up 2
- [[2026-07-03-cxmt-dram-capacity-memory-oligopoly-risk]] — CXMT, the ~77% intensity outlier in the Futurum comparison
Sources
~/rdco-vault/06-reference/research/2026-07-21-memory-maker-fab-financing-supply-side.md— [[2026-07-21-memory-maker-fab-financing-supply-side]] — both conflicting figures, verbatim~/rdco-vault/06-reference/research/2026-09-06-vendor-capex-composition-memory-anchor.md— [[2026-09-06-vendor-capex-composition-memory-anchor]]~/rdco-vault/06-reference/research/2026-09-07-memory-supplier-inventory-days-dio-anchor.md— [[2026-09-07-memory-supplier-inventory-days-dio-anchor]]- SK hynix Inc., Form 424B4, CIK 0002120882, filed 2026-07-10 — Tier A, primary: the mid-30% capex-to-sales policy sentence (p.101); audited consolidated cash-flow capex FY2023-FY2025 and Q1-2026; separate intangible-asset line; Note 12 PP&E additions by class; Note 24 DRAM/NAND/Others revenue split; single-reportable-segment statement; Solidigm consolidation; Note 29 capital commitments; Use of Proceeds (W45.5T facilities + W11.9T EUV, net proceeds US$26.2B); IFRS-as-issued-by-IASB basis; KRW presentation currency; W1,523.5/US$1.00 noon buying rate at 2026-03-31. Retrieval note: SEC EDGAR returns HTTP 403 to a browser User-Agent and requires a
User-Agentheader of the formName email. - SK hynix Announces FY25 Financial Results — SK hynix Newsroom, 2026-01-28 — Tier A: FY2024/FY2025/Q4-2025 revenue and operating profit
- SK hynix 2Q26 business results — SK hynix Newsroom, 2026-07-29 — Tier A: Q1-2026 and Q2-2026 revenue, net cash position, "CapEx Discipline" language
- SK hynix quarterly and annual cash-flow statement, KRX:000660 — stockanalysis.com, retrieved 2026-10-06 — Tier B: FY2021-FY2022 capex, Q2-2026 capex, D&A and operating cash flow; reproduces the audited lines to the won for every overlapping period
- SK hynix income statement, KRX:000660 — stockanalysis.com, retrieved 2026-10-06 — Tier B: FY2021-FY2022 revenue and the TTM column
- Will SK Hynix's Record $26.5bn ADR Issuance Help Close Its Capex Intensity Gap? — Brendan Burke, Futurum, 2026-07-10 — Tier C: the ~11% figure and its explicit "capital expenditure to revenue" definition; no underlying source cited for any of its four percentages
- SK hynix Expects to Maintain 2026 Capex at Mid-30% Revenue Level — Reuters via MarketScreener, 2026-01-28 — Tier C, body paywalled, headline and date only, not retried; superseded as evidence by the 424B4 policy sentence
- SK hynix post-Q profit beat offsets revenue miss — S&P Global Market Intelligence, 2026-08-05 — Tier B corroboration of the "high 40 trillion won range" FY2026 capex guidance, which post-dates the 424B4 and the Futurum piece