Picture a smoke detector wired to your position size. That is what the framework's oil rule is: when Brent crude closes above certain preset levels, the amount of money at risk gets cut automatically — no judgment call, no "let's see how it feels tomorrow." The first level, $83, tripped last Friday and halved the tactical book. The second level, $88.50, tripped on Tuesday: Brent ran as high as $89.31 during the session and finished around $89.25, its fifth straight up-day. By rule, the book drops to 30% of normal size, same session. (One honesty note, in keeping with house practice: the exchange's official settlement price couldn't be verified to the exact cent — one intraday reading said $88.23 — but three independent end-of-day sources cluster at $89.2–$89.3, all comfortably above the line.)
Why is oil doing this? The Strait of Hormuz — the channel carrying roughly a fifth of the world's oil — remains effectively closed, and the negotiation to reopen it has reached the point where both sides are demanding compensation from each other. Iran wants reparations, sanctions relief, and the naval blockade lifted; President Trump responded by demanding Iran pay compensation "for all the people it has killed." The US Navy has redirected 55 commercial vessels. Attacks in the Red Sea are disrupting the detour route. Goldman Sachs thinks Brent falls back to $80 by year-end if the strait fully reopens — the relief valve exists, but nobody is currently drafting the signature.
The bond market supplied the day's strangest footnote. The framework's rate rule arms when the 10-year yield closes above 4.70% two days running — Monday printed 4.71%, so Tuesday was the confirming close. The yield spent the whole day above the line, touching 4.74%... and settled at exactly 4.70%. Above the line is a trigger; on the line is not. The count resets — missed by literally zero basis points — but with CPI landing tonight, that reset could last all of one day.
Monday's flat close hid deterioration — 581 more falling stocks than rising ones. Tuesday's red close hid the opposite: 114 more stocks rose than fell, small-caps gained half a percent while the giants slipped, bank stocks rose again (191.16, less than a point from the level that would formally stand down the bear case's banking leg), and the Nasdaq's participation gauges recovered Monday's damage. The tape is trimming risk at the top of its range, not fleeing through the floor.
Not everything improved. The NYSE momentum oscillator spent a second day below zero, which under house rules upgrades last week's caution flag from "provisional" to confirmed. And the transport stocks — the economy's delivery vans — closed below Monday's low, a third straight decline while the big indexes sit near records. That divergence is old-school Dow Theory, and it's widening.
The bull path took the biggest cut — not because the evidence turned, but because the rules did: even a perfect inflation report tonight cannot re-risk the book past 30% while oil holds above $88.50. The bull case now needs oil to release before it needs stocks to rally. The shock path rose the most, because the oil leg went from "approaching the line" to "through it," with the next open air pocket running to the $95–$101 zone. The middle paths still dominate: a hedged, capped, coiled market chopping sideways while the verdicts grind out.
| Level | What it is | Why it matters |
|---|---|---|
| NDX 29,706 / 29,784 | Tuesday's and Monday's failed rally highs | Two probes at the ceiling died in two days — each one lower. Supply is winning the fight without the ceiling ever being touched. |
| NDX 29,427–29,341 | Tuesday's low, the gap floor, and the rising 50-day average | The first support stack — it held on Tuesday. A close below it opens the trapdoor. |
| NDX 29,109–28,777 | The open "thrust gap" from early August | The magnet below. If a hot CPI knocks the market down, this is where the elevator stops. |
| SPX 7,768 | The old upper Bollinger band | Tuesday opened exactly here and sold off all day — proven overhead supply. |
| SPX 7,629–7,600 | The open gap from last week | The designated buy zone if tonight's CPI knocks prices down — but only after the print, never before. |
| Brent $95 / $83 | The next tripwire up · the release line down | Above $95: book cuts to 20%. Back below $83: the cap comes off entirely. Oil is now steering equity position size. |
| 10-Yr 4.70% | The rate-rule line | Two closes above it re-arms the bond warning. Tonight's CPI can restart the count within hours. |
July CPI lands at 8:30pm Singapore time. Consensus: headline inflation 3.4% (down from 3.5%), core 2.5%. Markets price a September rate hike at roughly a coin-flip (~51%) — remarkable in itself, and the reason the stakes are asymmetric. A hot reading (core rising 0.3% or more on the month) prices the hike toward certainty, likely sends the market down through its gap ladder, and restarts the bond count the same night. A cool reading brings relief — but here's the rule change that matters: a cool print can no longer trigger buying. While oil sits above $88.50, the cap stays at 30% no matter how friendly the data. Cool means harvesting profits into strength, not chasing it. The rulebook has made the market defensive by construction, precisely so nobody has to be defensive by emotion at 8:31pm.
Producer prices follow Thursday, retail sales Friday. And underneath the week's noise, the long game stays quietly on script: the monthly momentum divergence — the three-peak structure across 2018, 2021, and now — is running at 69.87 versus its 77.88 cancellation line, retreating from the trendline rather than challenging it. Nothing this week has touched it.
Three items, per house practice. First: the standing order attached to the oil tripwire — "if Brent settles above $88.50, bank half the runner" — executed as written, locking gains of roughly 184 points on the banked half while the survivor position rides. Second: the rate-rule short that was one close from activating did not activate, because 4.70 is not above 4.70 — the framework does not round in its own favor. Third: a data discrepancy is logged rather than papered over — the put/call ratio we reported at 0.77 on Monday now shows revised to 0.61 in the source data (Tuesday printed 0.70). The conclusion — protection is bid — survives the revision; the magnitude doesn't. Both facts belong on the record.