Daily Macro US
pos-018 Debut
22.5c
Friday's conditional entry filled at 18c — +$6.30 in one session; crowd still at 2.0-2.5 (41c) vs GDPNow 1.7
Cleveland July Nowcast
3.32%
FELL from 3.71 — June actual reset the base; planned CPI wing trade killed by its own entry condition
July CPI Market
3.4 @ 37.5c
Crowd migrated to 3.4-3.5 on oil logic — model and crowd one bracket apart; no edge, no trade
USD/JPY 165 Touch
74.5c
Ran from 63c before rules verification — hedge window missed; latency lesson logged
Zero Cuts (pos-013)
84.6%
One point from original FV; compounder compounding
Week Ahead
Loaded
Japan CPI Wed → FOMC Jul 29 → Jul 30: BEA advance (pos-018) + BoJ statement (gate)
Three scan results today, and only one of them is a trade — which is the point. First, the trade that already happened: Friday evening's conditional entry executed itself. The automated check found GDPNow still at 1.7% and the US Q2 GDP 1.5-2.0% bracket offered at 18c (cheaper than the 24c we'd flagged that morning), so pos-018 went on at $25 — and opens this week at 22.5c, +$6.30 in a single trading day, while the crowd still parks 41c on the 2.0-2.5 bracket the model doesn't favor. Resolution July 30, BEA advance estimate. Second, the trade we planned and killed: Friday's plan was to buy the July-CPI 3.5/3.6 wing if Cleveland's nowcast confirmed at 3.6 or higher on the oil re-spike. It did the opposite — the July nowcast FELL to 3.32 (from 3.71), because the June actual (3.5) replaced the model's too-high 3.92 as its base, resetting the whole path lower, while its gasoline inputs haven't yet caught the newest Brent leg. Meanwhile the crowd migrated exactly where we'd planned to go: 3.4 to 37.5c, 3.5 to 22.5c. So the model now argues 3.3 and the crowd argues 3.4-3.5, one bracket apart, with 23 days of horizon error that swamps the gap — and the June lesson (the model misses when oil moves fast) cuts against BOTH readings. No trade. The scan that ends in a documented pass is the scan working. Third, the one that got away: the USD/JPY 165-touch hedge we flagged Friday at 63c — a natural pair against the BoJ gate — ran to 74.5c before we'd verified its resolution rules (Investing.com hourly candle highs; verified now, too late). At 74.5c the payout no longer hedges anything worth hedging. Logged as watchlist latency: rules verification is part of the entry, so it belongs in the same session as the flag, not the next one. The book otherwise idles green into a loaded week: zero-cuts at 84.6c one point from its original FV, the 10Y touch at 22.5c with both engines warm, pos-010 parked at fair awaiting an FOMC with $84 Brent on the table. Calendar: Japan CPI Wednesday, FOMC July 29, then the twin resolutions of July 30 — BEA advance (grades pos-018) and the BoJ statement (opens the two-week-old gate). Book: 6 open, $457 staked, +$21.13 unrealized, +$476.28 realized, 9/12.
Today's Market Moves
US GDP Q2 = 1.5-2.0% (pos-018 — NEW, Fri evening)
18%→22.5%+4pp
The conditional-entry system's first fully automated trade: condition checked (GDPNow 1.40-2.05, price ≤28c), filled at 18c, documented in Friday's post addendum. FV ~35 with the model-error haircut. One session later: +$6.30. The crowd's 41c on 2.0-2.5 is the other side of our bet, priced.
July CPI Annual (scanned, PASSED)
14%→22.5%+8pp
The 3.5 bracket we planned to buy at 14c trades 22.5c — but we're not kicking ourselves, because the entry condition (nowcast ≥3.6) failed in the opposite direction: Cleveland fell to 3.32. Model 3.3 vs crowd 3.4-3.5, 23 days out, with the model's known fast-oil blind spot on both sides of the argument. Pass documented; revisit when the nowcast stabilizes in early August.
USD/JPY Touches 165 (scanned, MISSED)
63%→74.5%+12pp
The BoJ-gate hedge ran 11.5pp over a weekend while its rules sat unverified. Rules are clean (hourly candle high ≥165, Investing.com source, any hour through Dec 31) — the process wasn't. New method note: rules verification happens in the same session as the flag, or the flag expires.
Zero Fed Cuts 2026 (pos-013)
83.9%→84.6%+1pp
Steady grind. 84.6c against FV 88; the position's whole job this week is to exist while the FOMC approaches.
Screening Table
| # | Market | Expiry | Market Price | Fair Value | Gap (pp) | Direction | Volume | Confidence |
|---|---|---|---|---|---|---|---|---|
| 1 | US GDP Q2 = 1.5-2.0% | Jul 30 | 22.5% | 35% | +13pp | HOLD $25 YES (entered Fri 18c) — GDPNow 1.7 vs crowd at 2.0-2.5 | $$15K | 5/10 |
| 2 | 10Y Touches 4.8% Before 2027 | Dec 31 | 22.5% | 38% | +16pp | HOLD $25 YES — dual engines warm | $$245K | 6/10 |
| 3 | BoJ 25bp Hike at Sep Meeting | Sep 2026 | 8% | 25% | +17pp | GATED — Japan CPI Wed, statement Jul 30-31 | $Minimal | 5/10 |
| 4 | Zero Fed Cuts 2026 | Dec 31 | 84.6% | 88% | +3pp | HOLD YES | $$12K/day | 7/10 |
| 5 | July CPI = 3.3% / 3.4% | Aug 12 | 26.5% | 30% | +4pp | PASSED — model/crowd one bracket apart, horizon error swamps; revisit August | $$16K | 4/10 |
| 6 | Fed Rate Hike 2026 | Dec 2026 | 52.5% | 56% | +4pp | HOLD YES — FOMC Jul 29 with $84 Brent | $$15K/day | 6/10 |
Market vs Fundamentals
Market Price (red) vs Estimated Fair Value (green) — %
Top 5 Opportunities
1
US GDP Q2 = 1.5-2.0% (advance) — YES
↑ BUY YES+13pp
Market price
22.5%
Fair value
35%
Gap: +13pp
The book's first fully automated conditional entry, and its early report card reads well: flagged at 24c Friday morning, condition-gated to the evening GDPNow refresh, filled at 18c, marked 22.5c Monday. The bet is simple — the model that mimics the BEA's own methodology says 1.7%, the crowd pays 41c for a bracket the model doesn't favor — and the sizing is humble because GDPNow's 0.77pp average error is the reason FV is 35 and not 60.
▵ Bull case
- GDPNow mimics BEA methodology — advance-estimate alignment is its home game
- Ten days to resolution, two more GDPNow refreshes to confirm
- Entry improved 6c by waiting for the condition — process alpha, documented
▿ Bear case
- 0.77pp model error spans three brackets — this is a plurality bet, not a certainty
- Trade-war-era inventory swings are GDPNow's known weakness
- Thin market: $38 of daily volume today
2
July CPI Annual (the pass) — YES
↑ BUY YES+4pp
Market price
26.5%
Fair value
30%
Gap: +4pp
Friday's plan met Monday's data and lost, correctly. We wanted the 3.5/3.6 wing on a rising nowcast; the nowcast fell to 3.32 instead (June's actual reset its base), while the crowd moved to exactly where we'd have bought. Now the model and the crowd disagree by one bracket with three weeks of error bars and the model's fast-oil blind spot arguing against trusting EITHER. The June loss taught us what this setup costs; the tuition stays paid.
▵ Bull case
- If the nowcast climbs back through 3.5 as gasoline catches Brent, the wing re-arms at better information
▿ Bear case
- One-bracket disagreements at 23-day horizons are noise dressed as signal
- The model just demonstrated a 0.42pp fast-oil miss in this exact direction-regime
3
BoJ 25bp Hike at September Meeting — YES
↑ BUY YES+17pp
Market price
8%
Fair value
25%
Gap: +17pp
Japan CPI Wednesday is the interim event; the statement July 30-31 is the gate. The hedge we'd designed against gate failure (USD/JPY 165 touch) ran away over the weekend, so the position — if entered — now runs unhedged; that raises the bar on what Wednesday's CPI and the statement language must show. The two-week discipline stands.
▵ Bull case
- Oil re-spike is imported inflation Japan can't intervene against
- 8c pays 12.5:1 post-gate
▿ Bear case
- Hedge window missed — gate failure now costs full downside
- Thin book unchanged
4
10Y Treasury Touches 4.8% Before 2027 — YES
↑ BUY YES+16pp
Market price
22.5%
Fair value
38%
Gap: +16pp
Recovered to 22.5c with both engines idling warm: the Fed engine waits on the FOMC, the war engine on every Iran headline. +$9 on the book's most asymmetric open position.
▵ Bull case
- FOMC with $84 Brent could reprice the whole curve
- Touch mechanics need one morning
▿ Bear case
- Ceasefire tail
- Correlation cap
5
Zero Fed Cuts 2026 — YES
↑ BUY YES+3pp
Market price
84.6%
Fair value
88%
Gap: +3pp
84.6 and climbing at walking pace. The FOMC on the 29th — hold or hike — is almost mechanically supportive: either outcome buries the cut narrative deeper. The position that asks nothing and pays anyway.
▵ Bull case
- Every macro path this week supports it
▿ Bear case
- 3pp left — watch for the capital-efficiency review at 90c+