We publish falsifiable markers before events and grade them in print afterward — hits, misses, and the uncomfortable ones in between. Today’s docket is crowded: a formal settlement note on J2, two rotation markers (F3, G3) that have quietly become the most important structural story in crypto ETFs, first cells on the K-series, and three L-series watches — all against a tape that just received its biggest dovish shock since Fed week: a canceled U.S. strike on Iran, an OPEC+ output hike, and crude down more than 4% overnight, per CNBC. Bitcoin trades between roughly $62,500 and $63,400 depending on provider.
The scorecard
| Marker | What we said | What happened | Grade |
|---|---|---|---|
| J2 (set Jul 30) | Jul 29 + Jul 30 flow cells sum negative | +$32.1M and +$233.1M — sum is +$265.2M | ❌ FAILED — formal settlement note below |
| F3 (set Jul 26) | Third straight week of ether funds out-drawing Bitcoin funds upgrades the rotation | Week ending Jul 24 was the third; the streak did not stop there | ✅ FIRED |
| G3 (set Jul 27) | Fourth consecutive week of ETH out-drawing BTC = trend, not anomaly | Week ending Jul 31: ETH +$27.4M vs BTC −$61.5M — fourth straight | ✅ FIRED |
| K1 (set Jul 31) | September hike odds ≥75% at Aug 7 close | Trackers split: low-60s Thursday flashes vs ~81% Friday prints; oil shock now argues lower | ⌛ TRACKING — divergence noted, final Aug 7 |
| K2 (set Jul 31) | MSTR closes every session in $85–$110 through Aug 7 | Friday close $93.97 (−3.86%) | ✅ HOLDING |
| K3 (set Jul 31) | Jul 31 + Aug 3 Farside cells sum positive | Jul 31 settled −$265.4M → today’s cell must exceed +$265.4M | ⌛ OPEN — deep underwater |
| L1 (set Aug 1) | Consolidated Coldcard-theft coins move by Aug 8 | All 1,367 stolen BTC (~$88.6M) remain 100% unspent per Galaxy | ⌛ TRACKING — unmoved |
| L2 (set Aug 1) | Any daily close above $65,000 by Aug 7 | No close above $65K yet; tape sits ~$63K | ⌛ NOT FIRED — four sessions left |
| L3 (set Aug 1) | IBIT prints a +$100M inflow day during Aug 3–7 | Window opens with today’s cell | ⌛ OPEN |
J2, formally settled: the cluster ended before we would let ourselves say so
J2 asked whether the July 29 and July 30 flow cells would sum negative — a bet that the four-session, $526.5 million outflow cluster would swallow the Fed decision itself. It did not: +$32.1 million on decision day, +$233.1 million the day after. We noted on Saturday that the marker was arithmetically dead; per our settled-cells-only rule, today is the scheduled formal grade: J2 FAILED. The instructive part is the shape of the miss. The cluster did not fade — it snapped, hard, the moment the decision removed uncertainty, and the buyer who returned (IBIT, +$89.8M and +$183M on consecutive days) was the same fund that had been the lone seller. Single-allocator flows around binary events look like trends and settle like coin flips. That lesson is now priced into how we build flow markers: calendar windows around events, not through them.
F3 and G3: the rotation is now the trend — and it fired on both definitions
Two separate marker series, set a day apart, asked the same structural question at different thresholds: F3 (July 26) wanted a third straight week of ether ETFs out-drawing Bitcoin ETFs; G3 (July 27) wanted a fourth. Both fired. The week ending July 31 saw ether funds add +$27.42 million while Bitcoin funds bled −$61.53 million, extending ether’s weekly winning streak to four, per The Crypto Times. The monthly totals are more lopsided still: ETH +$365.2 million in July against BTC +$172.4 million — and Bitcoin’s figure only survived above zero because the month ended before another outflow day. On Friday itself the split was −$265.4M BTC / +$9M ETH, per KuCoin.
Scale makes the signal stronger, not weaker. Bitcoin ETFs hold about $76.2 billion in assets against ether’s $9.7 billion — a seven-to-one ratio, per Cryptonews — so July’s flows represent roughly 3.8% of ETH complex assets arriving in a month while the BTC complex added 0.2%. Concentration cuts the other way, though: BlackRock’s ETHA accounted for 37,424 of the week’s 37,959 net ETH inflow — effectively the entire category through a single fund. We have seen what single-fund dependence does to Bitcoin flows in the IBIT era; graders of the next ETH marker should remember the rotation currently has one engine. Marker discipline note: fired markers retire. The rotation thesis now needs a reversal test, not another confirmation test — see O-series below.
K1 and the odds problem: grading into a fog bank
K1 asks where September hike odds close on Friday, threshold 75%. The honest status is that the input data disagrees with itself. Thursday post-FOMC flashes put CME FedWatch at 61.4% and 63.2%; Friday-morning prints reached ~81% (CME) and 82% (Yahoo). That 20-point spread across 24 hours exceeds anything we saw even in July’s divergence, and we flag it rather than average it. What we can say directionally: every input that pushed odds up in July — Hormuz risk, $100 Brent, the war premium — reversed over the weekend. A canceled “biggest attack since World War II,” a completed OPEC+ cut rollback, and a 4% overnight crude slide is the strongest single-session dovish package since the conflict began. If odds do not fall materially this week, K1 fires and tells us the hike case has broadened beyond oil (dissents, positioning, tariff pass-through). If they collapse, K1 fails and the barrel-not-spreadsheet regime we codified on July 31 gets its first confirmation from the dovish side. Either outcome is informative; that is what markers are for.
K2, L1, L2: the quiet ledger
K2 holds: MSTR closed Friday at $93.97, down 3.86% but inside the $85–$110 band for every session since the marker was set — the equity market continues to price the Digital Credit Capital Framework rather than the record $8.22 billion loss, even as CoinGape reports Saylor’s public focus shifting from Bitcoin accumulation to defending the 12% STRC dividend. Weekend reporting confirmed another week without Bitcoin purchases. L1 remains the tape’s strangest fact: all 1,367.05 stolen Coldcard BTC (~$88.6 million) across three attack waves remain 100% unspent, per Galaxy Research data relayed by AMBCrypto — dormancy that continues to defuse the forced-selling scenario. L2 (a $65K daily close by Thursday) needs roughly +3% from here with four sessions left; the oil shock is the first catalyst this week that could plausibly deliver it. L3 and K3 both hang on this week’s cells — K3 needing an implausible +$265.4 million today, a bar we set deliberately steep to test whether corporate-treasury selling and ETF demand are truly decoupled.
New markers: the O-series
With F and G retired and J settled, we open the O-series, grading by Wednesday, August 12 (CPI day):
- O1 — transmission test: CME FedWatch September hike odds print below 60% at any daily close by Friday, August 7. Fires = the oil de-escalation transmits to policy pricing at full speed. Fails = the hike case has decoupled from crude.
- O2 — seasonality floor: Bitcoin holds $60,000 on every daily close through the August 12 CPI print. August is the calendar’s only negative-median month; a held floor against four red Augusts of precedent would be the strongest counter-seasonal signal of the ETF era.
- O3 — premium unwind: Brent settles below $85 before the August 12 CPI print. Fires = the war premium fully drains and July’s CPI shock arrives already stale. Fails = supply risk is stickier than the headlines.
What would change our mind: if odds hold near 80% through a falling oil tape, our oil-first model of 2026 Fed pricing is wrong and we will say so. If the ETH out-draw streak breaks the week it gets crowned a trend, we will grade the irony as cheerfully as the streak. Open ledger: K1/K2/L2/L3 finals Friday, L1 Saturday, O-series and K3’s slow arithmetic through August 12, the ~August 11 difficulty retarget testing July’s floor thesis, and H1 — Saylor’s “another color” — with a week left on the clock.
FAQ
What is marker grading?
We publish specific, falsifiable predictions (markers) with explicit thresholds and deadlines, then report the outcomes in print — including the misses. It keeps our analysis accountable and auditable.
Why does the ETH-over-BTC ETF rotation matter?
Four consecutive weeks of ether funds out-drawing Bitcoin funds — $365.2M vs $172.4M for July — suggests incremental institutional crypto capital is choosing ETH exposure. With ETH complex assets one-seventh of Bitcoin’s, the relative impact is larger than the raw dollars.
Is the rotation broad-based?
No — and that is the caveat. BlackRock’s ETHA accounted for essentially the entire ETH category inflow last week (37,424 of 37,959 ETH). A one-fund trend is a fragile trend.
Why are rate-odds trackers showing different numbers?
Different snapshots, different methodologies, and a fast-moving tape: post-FOMC flashes showed low-60s while Friday prints hit ~81%. When trackers diverge we cite the range with dates rather than pick a side.
What are the key dates ahead?
August 7: payrolls plus K1/K2/L2/L3 finals. August 8: L1 final. August 10: H1 window closes. ~August 11: difficulty retarget. August 12: CPI and O-series finals.
Investment disclaimer: This article is for informational and educational purposes only and does not constitute investment, financial, legal, or tax advice. Bitcoin and cryptocurrencies are volatile assets; you can lose some or all of your capital. Always do your own research and consult a licensed financial advisor before making investment decisions.