ticker: 6146.T layer: L1-WFE verdict: red verify_date: 2026-05-28 #

Disco Corp (6146.T) — Graham Screen #

Screened: 2026-05-28
Source: yfinance (20-min delayed JPX data), Earnings dates from Wall Street consensus
FX assumption: JPY 160 = EUR 1.00

Snapshot #

YTD below 50% threshold — full screen triggered.

1. Business Overview #

Disco Corporation (founded 1937, TSE Prime) manufactures precision cutting (dicing saws), grinding, and polishing machines used in semiconductor and advanced packaging production. Holds ~80% global share in dicing saws and ~70% in wafer grinders — near-monopoly positions in critical backend semiconductor equipment. These tools are indispensable for HBM (High Bandwidth Memory) production and advanced 2.5D/3D packaging where CoWoS/SoIC adoption is accelerating. Also sells consumables (dicing blades, grinding wheels) with ~50% gross margins — a recurring revenue kicker.

2. Graham 7 Filters #

Score: 5/7

F3 note: FY2023 saw a trough quarter (JPY 117.09/qtr in Jul-2023 during WFE downturn) but no annual loss — stability confirmed. COVID quarters (2020) also above zero.

3. shr-017 Peak-Cycle Test #

This is the critical filter for Disco and the reason the supply chain map pre-flagged this as non-deep-value.

Quarterly EPS sequence (reported, JPY):

Annual EPS Trend:

Peak-cycle verdict: Current TTM EPS of 1,251 JPY is the highest in company history. Revenue grew 55% in 3 years. The stock re-rated +94% in 12 months, driven by HBM/advanced packaging narrative. This matches the Maersk/cyclical-trap pattern (shr-017) almost exactly — strong trailing score, but the question is forward guidance vs current pricing.

Critical forward signal: Next quarter (Jul 2026) analyst consensus = JPY 203.94, which is -48% QoQ from the Apr-2026 peak of 395.42. This is expected seasonality, but WFE downcycles in the past (2023) have surprised to the downside once orders decelerate. The US export control environment on advanced semiconductors to China (Disco's #2 market historically) is a structural wildcard.

4. Graham Intrinsic Value (shr-003/shr-012) #

Trailing EPS: JPY 1,251.16 #

Forward EPS: JPY 1,360.94 #

Break-even growth rate:

• Trailing EPS: 22.3% per year sustained for 7-10 years
• Forward EPS: 20.1% per year sustained for 7-10 years

shr-003 Implied growth gap:

• Trailing P/E 53x implies g = 22.3%
• Forward P/E 49x implies g = 20.1%
• Gap of only 2.1pp — the market is NOT pricing a profitability inflection. Both P/Es are already at premium "terminal value" pricing. There is no forward-multiple discount to buy.

Break-even assessment: The Graham formula requires Disco to compound EPS at 20-22% annually for a full decade to justify today's price. That means JPY 1,251 EPS growing to ~JPY 8,700-9,700 by FY2036. This assumes no WFE downcycle in a decade — historically impossible in semiconductor equipment. The 2023 trough was -40% from FY2022 EPS and came only 1 year after a cyclical peak. Demanding 20%+ CAGR through a full cycle with no misses is an extraordinary assumption.

5. Capital Allocation & Quality Metrics #

Balance sheet is genuinely exceptional — zero debt, JPY 285B cash. The moat is real. The issue is entirely valuation.

6. shr-020 Red Flag Inventory #

Red: 8 | Amber: 3 | Green: 2

7. Analyst Consensus #

20 analysts covering. Distribution: 2 Strong Buy, 10 Buy, 8 Hold, 1 Sell.

Bull case (100k): assumes HBM ramp continues 2-3 years at full pace and China exposure risk doesn't materialize.
Bear case (48k): ~35x forward P/E, implies multiple compression to "fair" WFE premium. Still above Graham value.

8. VWCE Overlap & Portfolio Fit #

Disco Corp is a constituent of MSCI ACWI and FTSE All-World, both of which underlie VWCE. Japan accounts for ~5.6% of VWCE weight; Disco is one of the top 30 companies on TSE Prime by market cap (~EUR 45B equivalent). The user already holds indirect Disco exposure through VWCE.

More importantly: VWCE also holds ASML, TSM, and has exposure to NVDA/AVGO via US tech weighting. Adding Disco as a satellite position would increase concentration in the AI infrastructure CapEx cycle — exactly the duplication the screening criteria flags. This is the same layer (equipment suppliers benefiting from AI CapEx) as ASML, just on the backend/packaging side rather than lithography.

9. Thesis Assessment #

The thesis is real. The timing and price are wrong.

Disco's competitive position is genuine: 70-80% market share in dicing saws and wafer grinders is not replicated by any competitor. Consumables provide recurring revenue. The HBM supply chain is structurally dependent on Disco's machines for advanced packaging. This is L1 supply chain (direct equipment layer), not a downstream play.

The problem: the market already knows all of this. The stock has re-rated from a trough of JPY 31,890 (12 months ago) to JPY 66,350 — fully reflecting the HBM ramp narrative. At 53x trailing P/E and 49x forward P/E, the market is pricing 20%+ sustained EPS growth for a decade in a notoriously cyclical industry where the last trough was a -40% EPS drop 18 months ago.

The supply chain map annotation was correct: "NOT deep value — recheck post-correction only."

10. Verdict #

VERDICT: RED

Disco Corp is one of the highest-quality semiconductor equipment businesses globally. The moat, balance sheet, and technology leadership are all exceptional. At today's price, none of that matters for a Graham value entry — the stock requires 20%+ annual EPS growth for a decade to break even on Graham's formula, with zero margin of safety at any realistic growth scenario.

This is a textbook peak-cycle pricing trap (shr-017). The WFE equipment sector historically corrects 40-70% from peak-cycle multiples — AMAT, LRCX, KLAC, and Disco itself all experienced this in 2022-2023. The next sector trough could put Disco at JPY 35,000-45,000 even without fundamental deterioration.

Additional disqualifiers for this portfolio:

• FCF yield of 0.74% — capital provides no income
• Dividend yield of 0.75% — no income cushion during drawdown
• Sector concentration duplicates VWCE's ASML/TSM holdings
• US export controls on China chip equipment are an unresolved overhang

Watchlist Trigger Prices #

Graham entry requires P/E compression to the 15-25x range on forward EPS. These are the trigger levels for post-correction recheck:

Practical watchlist trigger: JPY 32,000-35,000 (25-26x forward P/E on trough-cycle EPS). This range:

1. Assumes a ~50% price correction from current levels — historically consistent with WFE sector drawdowns
2. Assumes some EPS trough pressure (forward EPS falling from 1,361 toward 1,100-1,200 in a down-cycle)
3. At JPY 32,000 on EPS of 1,100 = 29x forward P/E — still a premium, but justifiable for the monopoly moat

Note: at those prices, VWCE re-evaluation is still required — a 50% correction in Disco likely means a broader tech/WFE downturn affecting the core VWCE holdings too.

Add pm watch add 6146.T below 35000 when/if there is a mechanism to monitor JPX tickers.

VERDICT LINE: RED — Disco Corp is a genuine monopoly in WFE backend equipment but priced at 53x trailing / 49x forward P/E with a 20%+ break-even growth requirement. Graham IV overvaluation is 65-84% at every realistic growth rate. Peak-cycle WFE pricing with historical 40-70% correction risk. Zero income yield. Sector duplicates VWCE. Recheck at JPY 32,000-35,000 (25-26x forward P/E trough).