Markers Friday: Oil Stopped Two Dollars Short, the Flow Test Failed, and Q2 Dies Tonight — Grading a Week of Pre-Commitments

Every Friday we grade in public. The markers below were committed in print before the data landed — R1 before Wednesday's CPI, R2 before Monday's flow prints, R3 before oil's Hormuz squeeze peaked, Q2 at the start of August. No adjustments after the fact, no quiet deletions. Here is how the week actually scored, and what survives into next week.

R1 — the September hike marker: DEAD, formally closed

R1 asked whether the September hike case could revive through the inflation data: it needed either a hot CPI (headline ≥3.5% or core >2.6%) or a hot PPI to push hike odds back above 50%. Neither leg fired. Wednesday's CPI printed exactly on consensus — +0.1% month over month, 3.4% year over year, core +0.2% and 2.5% — every figure on the Dow Jones number. Thursday's PPI then buried it: headline producer prices were unchanged in July against a +0.2% consensus, with final-demand goods down 0.7% and the year-over-year rate cooling to 4.7% from June's 5.5%.

Post-PPI, prediction markets price a September hold at roughly 60–71% (CME FedWatch via Motley Fool; Kalshi at 65%). The hike case that peaked at 67% on July 31 has been dismantled in two weeks by a negative payrolls print and two cool inflation reports. R1 is graded DEAD — and honestly, the in-line scenario we published Wednesday morning (hold ~60%, BTC pinned $62,700–65,200) is the one that played out, tick for tick.

One caveat we flag rather than hide: inside the flat PPI, portfolio management services — a direct PCE input — jumped 6.5%. If August 26's PCE runs hot off that component, the September conversation reopens. That is a new watch item, not a resurrection of R1.

R2 — the flow test: FAIL, closed and restated

R2 was the hidden-seller test: after the +$853 million inflow week reported by CoinDesk on August 9, we committed that a genuine bid would show ≥+$400 million across the Monday–Tuesday window. It printed roughly −$38 million (Monday −$144.67M, Tuesday +$4.89M, per Farside data). FAIL, formally closed on Thursday and restated here. Wednesday added another −$61.1 million (FBTC −$46.8M, IBIT −$14.3M, per FinanceFeeds), so the failure was not a two-day fluke — the bid simply left.

The grading nuance matters: the hidden-seller thesis itself scored UNTESTED, not falsified. A flow test that fails for lack of inflows tells you the buyer disappeared; it cannot tell you whether a large seller was absorbing the bid. The thesis is carried at unchanged confidence and retests the next time the complex strings together green days.

R3 — Brent $90 before August 14: FAIL, by two dollars

R3 was the week's live suspense: with the Strait of Hormuz still closed and Washington claiming "total control," would Brent touch $90 before today? It did not. The contract peaked at $88.91 at Tuesday's settlement (August 11), after a six-session, roughly 12% advance — then snapped, falling below $87 on Thursday as attention rotated to demand weakness (Trading Economics; Bloomberg via our Thursday coverage; Al Jazeera on the stalled reopening). Two dollars short. FAIL.

The reading is more interesting than the grade. A closed Hormuz, active attacks on shipping, and an IEA-estimated 1.8 million bpd quarterly shortfall could not push Brent through $90 — because the market is pricing the war as a rates event against a deteriorating demand backdrop, the same regime we documented in July. For Bitcoin, the transmission channel matters: the only realistic path back to a September hike ran through energy, and this week the energy market declined to build it. Geopolitics stays a rates story, and the rates story stays a hold.

Q2 — a daily close above $65,000: failing, final session today

Q2 was August's price quality marker: one daily close above $65,000 during the window that ends tonight. The high-water mark remains Monday August 10's intraday tag of $65,317 — a tag, not a close. Since then: $63,500 Wednesday, $63,766 Thursday morning, $63,283 at press time this morning. Barring a 3% Friday rally into the close — through an SEC rulemaking vote and a retail sales print — Q2 settles as a FAIL tonight. We will grade it formally Monday with the weekend candles on the board.

What a 0-for-4 week actually means

A week where every marker fails is not a failed week for the framework — it is the framework doing its job. Each grade encodes real information: the hike case is dead (R1), the ETF bid is absent (R2), the war premium is capped (R3), and the price cannot escape its range (Q2). Assemble those four facts and you get a coherent regime: a market with no macro catalyst, no flow catalyst, and no energy catalyst, drifting sideways at $63K while the biggest structural story of the year — the regulatory one — plays out in Washington rather than on the tape.

That is also why the failures cluster. R2 fails because R1's resolution (a hold, but a grudging one) gives allocators no urgency; R3 fails because the same demand weakness that killed the hike case caps oil; Q2 fails because without R2's flows there is no marginal buyer to force a $65K close. The markers are not four independent bets — they are four windows onto one stalled regime.

The Q-series review: what the ledger says after six weeks

Q2's likely failure tonight also closes the books on the Q-series' second test, so a brief review of the ledger is owed. The Q-series exists to grade price quality rather than price direction: instead of predicting where Bitcoin goes, each marker asks whether the market can do a specific, verifiable thing — hold a level on a weekly close, print a daily close through a ceiling — within a fixed window. Q1 passed in early August when the market defended its floor through the payrolls shock; it was graded PASS on August 11 with the caveat that defense is the cheaper half of the job. Q2 asked for the expensive half — one close above $65,000 — and the market has now spent the entire window demonstrating that the bid required to do it does not exist.

Two lessons carry forward into whatever Q3 becomes. First, the asymmetry itself is the finding: a market that can defend $62,000 through a war, a negative jobs print, and a hike scare — but cannot buy a 3% breakout through $65,000 in six weeks — is describing its own ownership structure. The holders are strong; the buyers are absent. That is exactly the regime the ETF flow data (R2) shows from the other side. Second, window length matters: Q2's month-long window turned out to be generous enough that its failure is meaningful rather than noisy. A marker that fails with room to spare is worth more analytically than one that fails at the buzzer, and this one will have failed with five days of shrugs to spare unless today surprises.

New markers for the week ahead

S1 (grades tonight): the SEC votes to publish Regulation Crypto, and we commit to the base case — unanimous 3-0, no Uyeda dissent. S2 (grades on publication of the text): the comment period is set at 60 days or fewer, keeping a pre-Peirce-departure final rule mathematically alive. C1 (grades September 15): CLARITY fails cloture — consistent with Galaxy's 30% and Polymarket's ~17% passage odds; if it instead advances, the entire Regulation Crypto timeline becomes a footnote, and we will say so. P1 (grades August 26): core PCE prints ≥0.3% month over month on the portfolio-management feed-through, reopening the September debate. Four commitments, on the record, before the events. See you at the grades.