Framework Read · The Journal
Shin-Etsu Chemical (4063.T): Markup Near the Highs, a Zero From the State Model, and a 47% Scar in the Record
Titan Macro Desk • 5 July 2026 • First entry in the 4063.T journal — every future update appends below, dated, never edited
Shin-Etsu, the world’s reference maker of silicon wafers and PVC, trades at ¥7,188, within 10% of its 52-week high after a 48.6% year, and the framework reads Markup. The quantitative state model prices the probability that this is a genuine bull state at exactly zero, and has filed the last 13 days as Sideways with an 11.5% bear probability attached. Somewhere behind both labels sits the number that explains the machine’s caution: a 47.1% maximum drawdown in this stock’s two-year record, the deepest scar in our Japanese coverage. Markets remember falls like that even when the crowd does not, and so does the model that lived through it. That argument, momentum against memory, is the entry.
The Investor Read: What Season Is This Stock In?
| Phase | MARKUP — trending higher, buyers in control |
| Quantitative state | SIDEWAYS — 13 days, bull probability zero, an 11.5% bear probability attached |
| Price | ¥7,188 (Japanese yen) — a ¥13.4 trillion market value, 9% below the 52-week high of ¥7,930 |
| Valuation | Trailing P/E 28.5 compressing to 21.8 forward — the earnings are arriving |
| Ethical screen | PASS, 70 — clears every screen, with caveats rather than headroom |
| Character | Moves about 2.7% on a typical day — and the record holds that 47.1% drawdown |
The season reads mid-summer with a specific, checkable engine: the multiple is compressing the right way, 28.5 trailing to 21.8 forward, which means earnings are growing into the price rather than the price outrunning the earnings. That is the healthiest valuation shape in this batch, and it separates Shin-Etsu from the premium names around it. The tape agrees: up 48.6% over the year to our June cycle, up 39.7% over six months, and the price has added another 7% since the June marker at ¥6,695. Against all that stands the state model’s flat zero on the bull question. When the same model held Nvidia at zero for 262 days through a rising tape, we called it a trend ageing in public. Here the tape is younger and the earnings shape better, so we weight the zero as caution rather than verdict, and we say so explicitly so this page can be marked against that choice.
The Trader Read: What Does the Tape Look Like Now?
Tactically the map has three lines. The June marker at ¥6,695 is the trend’s proof of work; the tape reclaimed and extended from it, and the recovery holds while it does. The 52-week high at ¥7,930 is the promotion line, 10% up. And below everything sits the memory of the 47% drawdown, which is not a level but a temperament: when this stock breaks, it breaks properly, with a 1.08 beta that keeps it tied to the global cycle argument in a way most of the Japanese wing is not. A 2.7% daily character gives the tape enough pace to resolve the state model’s doubt within weeks. The tactical read updates in the daily sessions.
Where the two reads stand: split, phase bullish against a state model that gives the bull case literally zero, with the tape siding with the phase layer so far. We flagged the same split at UPL and Nvidia in this journal; this is the version with the best earnings shape underneath it, which is why our lean here is with the season. Dated, so the lean can be scored.
The Tension: The Machine Remembers What the Crowd Forgot
The strongest fact against the bullish season is the record itself. A 47.1% maximum drawdown inside two years whose total return was just 22.9% means this stock has already once taken back everything it gave and more, and the state model’s refusal to certify the current run is partly that memory speaking through the arithmetic. Cyclical chemistry and wafer pricing are the fundamental version of the same warning: the 21.8 forward multiple is a promise about a cycle, and cycles in this industry have never yet kept all their promises. The insider and political files are empty, logged as empty. The street’s median target of ¥7,930 sits exactly on the 52-week high, a target that says continuation, not re-rating. If the zero-probability bull reading persists while the price stalls under ¥7,930, this page’s lean toward the season will have been the wrong call, and the appended entry will say so in those words.
What Would Change the Read
- The state model: the bull probability lifting off zero validates the season lean; the 11.5% bear probability growing instead is the earliest warning this page will get.
- The high: a weekly close through ¥7,930 takes out the street’s own finish line and forces targets, and this entry, to be rewritten upward by appendix.
- The June marker: losing ¥6,695 unwinds the extension that justified the Markup label, and the label goes to review.
- The earnings shape: the 28.5-to-21.8 compression is the bull case. Forward estimates slipping would remove it, and with it our stated lean.
Journal — first entry
5 July 2026 — ¥7,188 — MARKUP (state model: sideways, 13 days, bull probability zero). Journal opened on momentum against memory: a 48.6% year and the healthiest multiple compression in the batch, against a state model at zero and a 47.1% drawdown scar. Our lean: with the season, stated so it can be scored. The lines: ¥6,695 below, ¥7,930 above. Next review: the state moving, a line breaking, or estimates slipping, whichever is first. This entry is permanent.
Titan Macro Desk. This is analysis and education, not financial advice. Markets carry risk. Always manage your position size and do your own research.