For three issues we wrote that bitcoin was not responding to anything. On Friday it responded.

July payrolls printed at negative 23,000 against a street consensus near 80,000. Not a miss — a contraction, and the first monthly decline in months. Bitcoin took the number and ran, tagging roughly $65,340 intraday for its August high, closing the week near $64,900 and holding there through the weekend.

The repricing was violent. Polymarket's September 25 basis point hike, which we quoted at 57.5% a week ago and 49% on Wednesday, now trades at 35.5%. No-change has taken over as the base case at 62.5%, on more than $20 million of volume. On Wednesday we argued that the hike reprice was structurally premature because no data supported it. Two days later the data arrived and settled the question in one print. We were right about the mechanism and wrong about the direction it would break.

Here is the complication, and it is the whole issue. The unemployment rate did not rise. It fell — to 4.1% from 4.2% — and it fell because the labor force shrank, not because anyone was hired. Average hourly earnings growth slipped to 3.2% year over year, the softest since May 2021. Government shed 53,000, with local government education alone down 50,000. Retail lost 19,000.

So: payrolls went negative, wage growth hit a five-year low, the participation base eroded, and the headline rate improved. That combination is not a soft landing. It is a labor market deteriorating underneath a number that looks fine, and the number that looks fine is the one that leads the coverage.

Markets read the headline and bought risk. We think the composition is worse than the print. The FILTER makes that case.

Three metrics. Here is the read.

Metric 1 — Fear & Greed: 31. Still fear, but it finally moved. This index sat at 27 to 28 for a fortnight — frozen through the Coldcard sweep, the Strategy sale, the Fed, and a record day in equities. Four points is not euphoria and it is not a regime change. It is a crowd being shown something, which is exactly what we said it was waiting for.

Metric 2 — ETF Flows: +$853.5 million for the week. The largest weekly total since mid-April, and the answer to the question we left open. On Wednesday we said one green day is not a streak and told you to check. It became five: $170.1M Monday, $211.5M Tuesday, $244.4M Wednesday, roughly $102M Friday. BlackRock's IBIT took $693 million of it — 81 cents of every dollar. The bid is real, and it is also more concentrated in one issuer than at any point we have tracked.

Metric 3 — BTC Dominance: holding above 56%. Essentially unchanged from the 56.6% we quoted Wednesday, which means the rally did not rotate — bitcoin led and the rest of crypto came along rather than the reverse. Meanwhile Strategy's STRC closed Friday at $95.01, up from $92.32 on Wednesday. The gap to the $99–100 par that Saylor has stated is the buyback trigger has narrowed from roughly seven dollars to five in three sessions. That gap is the switch on 842,138 coins.

A dated test resolved on Friday, and it resolved messily. The full board is in the POLYMARKET STACK.

📣 THIS SPOT IS OPEN

Baseline Crypto reaches 7,000+ Bitcoin-focused readers every Monday and Wednesday morning. Direct sponsor placements are open — premium position, top of the issue, one per send.

Reply to this email for rates.

Concede the first half completely, including the part that cost us. On Wednesday we argued the September hike reprice was structurally premature because no release supported it — ISM had printed 55.6, JOLTS was in line, and the market was fading a hike on tone. Two days later payrolls came in at negative 23,000 and the reprice stopped being premature. Hike odds went from 49% to 35.5%. We had the mechanism right and the direction wrong, and the honest version of that is not a win.

Now the category error. The market has moved from “hike” to “hold,” not from “hike” to “ease.” September no-change trades at 62.5%. A 25 basis point cut trades at 1.9%. Fifty or more prices at under one percent. There is no cut in this market at any horizon that matters, and treating a collapsed hike as a dovish pivot is reading the absence of tightening as the presence of easing.

The composition is the second problem, and it is the bigger one. The unemployment rate fell to 4.1% while payrolls went negative. Those two facts only reconcile one way: the labor force shrank. People stopped looking. Add average hourly earnings at 3.2% year over year — the softest since May 2021 — and the picture is not an economy cooling gently toward a Fed that can relax. It is one where the headline unemployment rate is actively concealing the deterioration underneath it, which is the configuration in which central banks are historically late rather than early.

Then Wednesday. July CPI prints at 8:30 AM Eastern with consensus near 2.8% headline against 2.7% in June, and core around 3.0%. If core comes in at or above 3.0% while payrolls are negative, Warsh's Fed is looking at the one configuration its stated reaction function does not resolve cleanly: inflation that is not falling and a labor market that is. A hawkish reaction function plus a weakening labor market does not produce a hold by default. It produces a fight.

Dated test — and read the grading date carefully.

Branch A — the hold holds. Core CPI prints at or below 2.9%, and September no-change is still above 60% at Friday's close. Read: the labor turn dominates, September is effectively decided, and the market spent last week correctly pricing an economy that is slowing faster than the headline rate admits.

Branch B — inflation re-arms the hike. Core CPI prints at or above 3.1%, and September 25 basis point hike odds close Friday back above 45%. Read: Warsh gets the print that lets him tighten into a weakening labor market, the September question reopens, and the $61,400 support we have been watching since the start of the month matters again.

We grade this on Monday, August 17 — not Wednesday. CPI lands at 8:30 AM and Wednesday's issue goes out at 7:30, one hour earlier. We set a test we could not grade once already this month and we are not doing it twice. Wednesday gets the setup. Monday gets the verdict.

Three numbers that define the week ahead.

3.0% — core CPI, Wednesday at 8:30. Consensus sits near 2.8% on the headline against June's 2.7%, with core around 3.0%. This is the print that either closes the September question or reopens it, and it is the single number that decides the dated test above. Note the timing: it lands an hour after Wednesday's issue reaches you, so read our setup on Wednesday and check the number yourself at 8:30. If core surprises above 3.1% with payrolls already negative, the calm in this market gets tested immediately.

$95.01 — STRC, five dollars from the switch. Friday's close, up from $92.32 on Wednesday, against a $99–100 par that Saylor has publicly tied to both the buyback and the restart of bitcoin buying. Three sessions took two dollars off the gap. Watch this ticker the way you would watch an ETF flow print — it is the cleanest public signal for when the largest corporate holder stops being a seller. Saylor's stated arithmetic still points at roughly September 8.

2.53% — what a soft fork looks like when it dies. BIP-110's final signaling number, against a 55% threshold. The minority chain it produced lasted one block. Keep this figure somewhere retrievable, because the next contested proposal will arrive with confident claims about miner sentiment, and 2.53% is a useful reminder of the distance between loud advocacy and hash power. The eCash hard fork at block 964,000, around August 21, is the next scheduled test of the same question.

Forget analyst predictions. Polymarket is a real-money prediction market — traders put actual dollars on outcomes. Scorecard first, then the board. Not financial advice. Our read. Disclosure: we hold personal positions in Polymarket itself and may earn a commission from Polymarket referrals. These markets are thin and the odds below move — direction over ticks.

Scorecard: two legs of three, and the leg that failed is the interesting one. Wednesday's issue set a dated test with two branches and promised not to move the goalposts, so here it is against the tape. Branch A required three things. Payrolls at or below 70,000: passed, emphatically, at negative 23,000. September hike odds closing under 45% and staying there: passed, at 35.5% through the weekend. Unemployment ticking to 4.3% or higher: failed — it went the other way, to 4.1%. Branch B failed on every leg. So we do not get to call this a clean Branch A, and we are not going to.

What we got right and what we got wrong. The underlying claim across two issues was that the next repricing would come from labor rather than CPI, and that is exactly what happened — one labor print moved September from 49% to 35.5% with no new inflation data. What we got wrong was assuming the labor turn would show up in the unemployment rate. It showed up in participation instead, which is the more troubling place for it to appear and the reason the FILTER above is more bearish than the tape. Direction right, indicator wrong, no credit claimed on the branch.

No position was taken into the print. Wednesday's issue said no trade until Friday's payrolls and the book stayed flat. Given that the contract moved thirteen points on a number nobody forecast, standing aside was worth more than the trade would have been.

Fed decision in September — 25 bps increase | Market: 35.5% ⚪ NO TRADE — the test resolved, the edge did not
Down from 49% Wednesday and 57.5% a week ago, with no-change now the base case at 62.5% and cuts of any size under 3% combined, on more than $20 million of volume. The dislocation we flagged against CME futures has closed from the Polymarket side rather than the futures side. We do not chase a contract that has already made its move; the next entry, if there is one, comes from Wednesday's CPI, not from Friday's payrolls.

Zero Fed rate cuts in all of 2026 | Market: 85.75% ⚫ EXITED — and the exit was early
We shorted this at 78%, it went against us, and we covered at 88.75% on a pre-stated rule. It now trades at 85.75%. The payroll miss did crack the no-cut consensus by three points, which means the thesis was not wrong so much as early, and the rule that protected us from a larger drawdown also took us out before the turn. Both of those things are true and we are not going to pretend the second one away. Still a realized loss. Not re-entering on a three-point drift.

Bitcoin dips to $55,000 before 2027 | Market: 56.5% 🟡 WATCHING — trigger did not fire and cannot now
Drifted up two points from Wednesday's 54.5% even as spot rallied, which tells you the market is pricing the drawdown risk independently of the last three sessions. Our published entry trigger required September hike odds above 65% while this traded under 60%. The hike leg went to 35.5% — the wrong direction by a distance — so the trigger is dead in its current form rather than merely unfired. We will restate it or retire it on Monday the 17th once CPI has resolved.

CLARITY Act signed into law in 2026 | Market: 20.5% ⚫ EXITED — rule keeps validating
Down another two points from Wednesday and roughly half the 41% where the position was first opened. We closed at 33–37% on the stated cloture rule on July 29, before Thune shelved it. Congress is in recess. Re-entry only if the merged text survives intact and the calendar reopens in September, and nothing about a drifting price changes that.

Track the whole board live at polymarket.com — free, no account required.

A reader called the outcome before we did. His mechanism never fired.

Two issues ago Darryl wrote in arguing there would be no hike this year, and got there through a channel our research had not considered: Trump needs the Strait of Hormuz open before the November primaries, oil falls, inflation falls, the Fed stands down. We ran it, said it was a cleaner explanation than anything in our own FILTER, and put it on the record before Friday rather than after.

Friday tested it. He was right about the destination. September hike odds have collapsed from 57.5% to 35.5% and no-change is now the base case, which is close to what he described.

The route was not his. Brent settled Friday at $83.55, up 1.3% on the week. The Strait is not open. Iran and Oman are still negotiating transit terms, with Iran seeking to exclude US and Israeli vessels and Washington pushing for a return to pre-war conditions. Oil did not fall, inflation was not dragged down by energy, and the hike died anyway — killed by a payrolls print, which is the channel our FILTER argued for two issues running.

That is worth stating plainly because it is the kind of thing that usually gets blurred. A correct call through a mechanism that did not activate is not the same as a correct thesis, and the difference matters for what you do next. Darryl's oil channel is still loaded and still unfired; if the Strait reopens into a Fed that has already stopped tightening, that is a different and more interesting trade than the one he described.

One question: now that labor has done the work, does Wednesday's CPI still matter — or was September decided at 8:30 on Friday morning?

Hit reply. We read every response, and the best calls run Wednesday.

See you Wednesday — with the setup into CPI, whether the no-change base case held through the weekend, and how much closer STRC has moved to par.

— The Baseline Crypto Team