Financial Shenanigans

Financial Shenanigans

Disco's reported numbers look like a faithful representation of economic reality. Across 26 years of standardized Japanese-GAAP financials, ten years of cash-flow conversion testing, and every category of the shenanigans taxonomy, we find no restatements, no auditor changes, no regulatory actions, no material weaknesses, no non-GAAP adjustments, no acquisition accounting, no factoring, and no supplier finance. The one yellow flag worth tracking is the FY2026 step-up in receivables — DSO drifted from 40 to 48 days while revenue grew 11.1% — but it is small in absolute terms and consistent with a faster-growing overseas mix (89.6% of sales). The single thing that would change the grade is a recurrence of that DSO drift into FY2027 without a matching cash-collection explanation.

1. The Forensic Verdict

Risk grade: Clean. Three-year CFO/Net Income averages 1.02; the FY2026 accrual ratio is 0.28%; non-operating income contributes essentially nothing (negative 0.03% in FY2026); goodwill and intangibles together are under 0.5% of total assets; and the MD-A reconciles "free cash flow" explicitly as CFO plus CFI rather than burying definitional gaps. The only items moving the needle are governance-disclosure opacity inherent to Japanese statutory framework and one quarter of receivable growth that warrants follow-up — neither is a thesis breaker.

Shenanigans scorecard — all 13 categories

Twelve of thirteen categories pass clean; one is yellow (revenue/receivables drift); one balance-sheet disclosure (cash and deposits ≠ cash and equivalents) is yellow on optics but green on substance because management reconciles the ¥100B time-deposit move explicitly in the MD-A.

2. Breeding Ground

Disco's structural setup is benign on most dimensions and modestly elevated on key-person concentration. The compensation policy explicitly avoids the kind of growth or stock-price targets that incentivise aggressive reporting.

The Sekiya founder-family CEO tenure (49 years) is the only structural lever that increases shenanigan-likelihood materially, but the compensation design — anchored to a conservative 4-year cumulative ordinary-margin target that the company has been double — neutralises the usual incentive vector. Disco's deliberate 1-quarter guidance window also reduces the analyst-expectation pressure that often drives aggressive period-end recognition.

3. Earnings Quality

Earnings quality is high. Margin expansion is supported by mix (HBM/AI logic equipment over commodity dicing) rather than by reserve releases, capitalization, or non-operating gains. The only test that flashes yellow is FY2026 receivables.

Notes plus trade receivables expanded 33.1% (¥43.2B → ¥57.5B) while revenue grew 11.1%, lengthening DSO by 8 days. Two non-fraud explanations carry most of this gap: (1) overseas-sales mix rose to 89.6% in FY2026 (from 87.8% in FY2024), and overseas equipment customers typically take longer to remit than Japanese customers; (2) the FY2026 acceptance cycle was loaded toward year-end logic/HBM shipments. We flag it because if AR growth persists into FY2027 without a remit catch-up, the cleanest forensic test on the income statement starts to weaken. The allowance for doubtful accounts grew ¥140M → ¥817M (5.8x) on a tiny base — directionally conservative, not aggressive.

The operating-margin expansion from 23.4% (FY2017) to 42.3% (FY2026) tracks mix-shift evidence in the segment data, not cost-capitalization. Capex/Revenue is procyclical (peaks at the top of capacity-build years and troughs in steady-state years), not a hidden cost shelter. There is no capitalized software, no capitalized contract acquisition cost, and no "other intangibles" growth — investments and other assets sat at ¥25.3B both years.

Non-operating income contributes essentially nothing to reported earnings, meaning operating income carries the full economic story. There is no gain-on-sale, equity-method, JV-uplift, or FX-gain pattern that would inflate sustainable earnings.

4. Cash Flow Quality

Cash flow looks earned, not engineered. Ten-year cumulative CFO/Net Income is 1.10, and the working-capital contribution to CFO in the most recent year was negative, not positive — meaning Disco's CFO grew while paying suppliers faster, which is the opposite of the classic "operating cash flow lifeline."

Notes and accounts payable plus electronically recorded operating obligations fell from ¥30.9B to ¥22.9B — a 26% drop during a year in which revenue grew 11% and COGS grew 13%. That is the opposite of the classic "stretch the vendors" CFO lifeline. CFO still grew 11% because of underlying earnings power, not working-capital pressure.

The FY2026 reported FCF is negative ¥2.2B, which looks alarming until you bridge it to the cash-management decision disclosed in the MD-A: ¥100B was moved from cash-equivalents into time deposits during the year. Mechanically that ¥100B flows through "investing" (purchase of time deposits) but is not an investment in the operating business. Adjusting for the time-deposit reclassification, "true" FCF for FY2026 is closer to ¥97.8B. The MD-A states this explicitly — there is no obfuscation. The presentation is transparent.

Capex/D-A averages 3.0x over the last decade. In isolation this looks like operating-cost capitalization, but cycle-by-cycle Disco builds capacity (FY2022, FY2025) and then digests (FY2023, FY2024). The pattern is procyclical, not smoothing. PPE growth is overwhelmingly in buildings, land, and machinery — visible, physical assets that show up in the disclosed BS sub-lines.

5. Metric Hygiene

Disco's external metric set is unusually clean by global standards: there is no adjusted EBITDA, no non-GAAP EPS, no "cash earnings," no "free cash flow ex-X" definition gymnastics. The metrics management highlights — ordinary income, RORA, equity ratio — all reconcile transparently to the financial statements.

The "shipment value" line that Disco discloses alongside "net sales" is a sector-standard durables-equipment KPI, not an opportunistic re-frame; both numbers reconcile to the same shipment schedule and are presented every quarter. The one bridge the reader must keep in mind is FCF: when management writes "free cash flow was negative ¥2.2 billion," they mean CFO + CFI, which includes the ¥100B time-deposit transfer that is not an investment in the operating business. Subtracting that, the operating FCF run-rate is approximately ¥97B — consistent with the 5-year average.

6. What to Underwrite Next

The forensic work moves the investment dial only marginally. Disco's accounting risk is a footnote, not a haircut, and not a position-sizing limiter.

Track these signals in the next four quarters:

(1) Receivable normalization. DSO of 48 days is not high in absolute terms, but the FY2026 step-up from 40 days deserves one quarter of follow-up. If Q2-Q3 FY2027 collections do not bring DSO back into the 40-45 range, the question becomes whether revenue is being pulled forward via more generous terms. If they do, the yellow flag goes green.

(2) Time-deposit treatment. The ¥100B reclassification was clean and disclosed this year. If it grows or migrates between "cash and deposits" and "cash and equivalents" without a written bridge in the next MD-A, the disclosure trust score declines.

(3) Capex re-engagement after FY2027. FY2025 capex ¥69.8B reset above depreciation. If FY2027 capex falls back below ¥30B without a stated reason — as happened in FY2023-FY2024 — the procyclical pattern continues; if it falls without a stated reason, the question of disinvestment becomes worth asking.

(4) CEO succession signal. Kazuma Sekiya has run the company since 1977. Public succession architecture is thin. A formal successor announcement or co-leadership disclosure would be governance-positive; absence into the late 2020s extends key-person risk.

(5) Allowance for doubtful accounts trend. The 5.8x build (¥140M → ¥817M) is tiny in absolute terms but directional. A continued build would signal that overseas customer credit deterioration is more than a one-quarter event.

What would downgrade the grade: Any of (a) DSO drifting above 55 days; (b) factoring or supplier-finance program newly disclosed; (c) adjusted EBITDA or non-GAAP EPS suddenly introduced; (d) auditor change or any emphasis-of-matter paragraph from KPMG; (e) related-party transactions surfacing in the EDINET Yuho.

What would upgrade the grade: Continued multi-year accrual ratio under 0.5%, DSO normalization, and a published written succession plan.