This column’s Sunday analysis set three markers for the week and promised public grades with no retroactive redefinitions. Two of the three are now settled, a day ahead of Friday’s formal scorecard, and the third is producing the only genuine suspense of the week. Here is the accounting — the misses printed as prominently as the hits, as ever.
R1: dead on leg one
R1 was deliberately two-legged: it fired only if core CPI printed ≥0.3% month over monthandSeptember hike odds exceeded 50% at Thursday’s close. Wednesday’s print killed it at the first gate — core came in at +0.2% on the month, 2.5% annually, in line with every forecast. The second leg never mattered: by late Wednesday morning FedWatch showed hold probability near 62%, and Bloomberg’s bond-desk survey had the hike wager parked around 40% — below the trigger, moving the wrong way. R1 does not fire. That is not a failed prediction; it is the marker doing its job. The scenario table we pre-committed on Sunday and restated Wednesday said an in-line print meant “status quo — Fed futures hold near 60% pause; bitcoin likely stays pinned in the $62,700–$65,200 range.” The tape delivered: hold odds 62%, bitcoin at $63,766 as of 06:10 UTC Thursday per CoinGecko, range untouched. The in-line scenario was the boring one, and it was also the one that happened. For the record, this was the pre-print snapshot from CME’s own rates desk:
#FedWatch: 50/50 probability between September rate hike and keeping the current rate. All eyes on CPI and PPI data release.
— CME Group Interest Rates (@Interest_Rates) August 12, 2026
One leg of the week’s data remains: July PPI, this morning at 8:30 a.m. ET. June printed −0.3% on the month with the annual rate at 5.5%; consensus calls for +0.2%, with prediction markets leaning toward the annual figure cooling to 5.1% or lower. A hot PPI surprise is the one release left that could push Thursday’s odds close back above 50% and force us to reopen a marker we have just pronounced dead — we flag the possibility precisely so nobody can accuse us of burying it. Our new guide to reading the report is here for the mechanics.
R2: FAIL, formally — and what it does to the hidden-seller thesis
R2 required the settled Farside cells for August 7 + 10 + 11 to sum to at least +$400 million. Final arithmetic: +$101.7M (Thursday) − $144.67M (Monday) + $4.89M (Tuesday, per FinanceFeeds) = roughly −$38 million. Against a $400 million bar, that is not a near miss; it is a category miss. R2: FAIL.
| Session | Net flow | Detail |
|---|---|---|
| Thu Aug 7 | +$101.7M | IBIT +$86.7M, FBTC +$41M; settled cell |
| Mon Aug 10 | −$144.67M | Five-day streak snapped; IBIT −$53.56M, GBTC −$52.02M |
| Tue Aug 11 | +$4.89M | IBIT +$50.19M vs FBTC −$40.32M — a wash |
| R2 total | ≈ −$38M | Bar: ≥+$400M → FAIL |
Now the honest part. R2 was designed as a forcing test for the hidden-seller thesis — the observation that last week’s $853.5 million of inflows absorbed roughly 5.9× new issuance while price failed to close above $65,000, implying somebody supplying coins into the bid at scale. The test’s logic: if flows stayed heavy and price still could not clear $65K, the thesis would graduate from suspicion to working model. Flows did not stay heavy — they collapsed to roughly zero. So the correct conclusion is not “thesis falsified” and certainly not “thesis confirmed”; it is thesis untested. Weak flows plus a weak price is exactly what an ordinary no-demand week looks like, and requires no hidden seller at all. The interesting version of the question — who sells into a strong bid? — returns only when the bid does. We carry the thesis forward at unchanged confidence, with this week logged as a null result. (The candidate list is unchanged too, and one candidate did something notable Wednesday: Metaplanet moved 9% of its treasury between wallets with its position $1.4 billion underwater — today’s update covers what that does and does not prove.)
R3: two dollars of suspense
R3 fires if Brent settles above $90 before Friday, August 14. When we armed it on Sunday, Brent was below $82 and the marker looked like a throwaway — we wrote “Far” next to it. Five sessions later it is the closest thing this week has to live ammunition: Brent ran 12% over six sessions to touch $88 before snapping the streak early Thursday, per Bloomberg, driven by attacks that dented Hormuz reopening hopes, President Trump’s claim of “total control” over the strait, and an IEA estimate of a 1.8 million barrel-per-day global shortfall this quarter. The marker’s design intent was exactly this configuration: an oil re-spike arriving after a tame CPI, threatening to reimport energy inflation just as the Fed talks itself into a hold. If Brent tags $90 today, the hike case revives on Friday’s odds board no matter what core CPI said; if it fades from $88, R3 joins R1 and R2 in the miss column and the week closes with the range — and the pause — intact. Either way it grades tomorrow, in public, against the number we wrote down on Sunday.
The scoreboard into Friday
| Marker | Trigger | Status as of Thursday 06:10 UTC |
|---|---|---|
| R1 — CPI/rates | Core ≥0.3% m/m AND hike odds >50% Thu close | DEAD — core printed 0.2%; odds ~40% and falling. Formal close at tonight’s odds settle; PPI is the only reopen risk |
| R2 — ETF bid | Aug 7+10+11 settled flows ≥ +$400M | FAIL — ≈−$38M. Hidden-seller thesis: untested, carried forward |
| R3 — Oil shock | Brent settles >$90 before Aug 14 | LIVE — $88 after +12% in six sessions; grades Friday |
| Q2 — price | One daily close >$65,000 by Friday | FAILING — two sessions left; Monday’s $65,317 tag still the high-water mark |
What would change our mind between now and Friday’s grades: a PPI print hot enough to drag hike odds back through 50% today (reopens R1 on its own terms — both legs, no fudging); a Brent settlement above $90 (fires R3 and stress-tests the 2026 correlation regime we have been mapping since the July Hormuz closure); or a $65,000 daily close (fills Q2 and forces us to retire the range-bound framing that has organized a month of coverage). Three specific numbers, two sessions, all checkable by anyone with a free data feed. That is the whole method: the column earns whatever trust it has on the weeks the numbers come back wrong and we print them anyway.
What were the three markers this week?
R1: core CPI ≥0.3% m/m plus September hike odds above 50% at Thursday’s close. R2: settled ETF flows for Aug 7+10+11 of at least +$400M. R3: a Brent settlement above $90 before August 14.
Why did R1 fail if the CPI matched forecasts?
R1 did not “fail” as a prediction — it did not fire. It was built to detect a hot-inflation regime change; an exactly in-line print is the definition of no regime change, and the pre-committed in-line scenario (range-bound bitcoin, ~60% hold odds) is what occurred.
Is the hidden-seller thesis dead?
No — untested. The test required heavy inflows against a capped price; inflows went to zero instead, which any ordinary soft-demand week would also produce. It gets retested the next time the bid returns.
What could still move bitcoin this week?
Today’s 8:30 a.m. ET PPI print, a Brent move through $90, Friday’s SEC Regulation Crypto vote, and Friday’s retail sales data.
Disclaimer: This article is for informational purposes only and does not constitute investment advice. Cryptocurrencies and crypto-linked equities are volatile and you can lose money. Do your own research and consult a licensed financial advisor before making investment decisions.